fix(chat): repair inconsistent quote escaping in edit_diagram JSON

When the LLM generates edit_diagram tool calls, it sometimes produces
inconsistent quote escaping in XML attributes within JSON strings.
For example: y="-20\" instead of y=\"-20\"

This causes JSON parsing to fail, and jsonrepair cannot fix this pattern.

Added pre-processing regex to detect and fix cases where the opening
quote is unescaped but the closing quote is escaped in attribute values.

.gitignore

@@ -68,4 +68,5 @@ CLAUDE.md
 
 # edgeone
 .edgeone
-opencode.json
+opencode.json
+ai-models.json

README.md

@@ -207,6 +207,7 @@ See the [Next.js deployment documentation](https://nextjs.org/docs/app/building-
 -   OpenAI
 -   Anthropic
 -   Google AI
+-   Google Vertex AI
 -   Azure OpenAI
 -   Ollama
 -   OpenRouter
@@ -221,6 +222,10 @@ All providers except AWS Bedrock and OpenRouter support custom endpoints.
 
 📖 **[Detailed Provider Configuration Guide](./docs/en/ai-providers.md)** - See setup instructions for each provider.
 
+### Server-Side Multi-Model Configuration
+
+Administrators can configure multiple server-side models that are available to all users without requiring personal API keys. Configure via `AI_MODELS_CONFIG` environment variable (JSON string) or `ai-models.json` file.
+
 **Model Requirements**: This task requires strong model capabilities for generating long-form text with strict formatting constraints (draw.io XML). Recommended models include Claude Sonnet 4.5, GPT-5.1, Gemini 3 Pro, and DeepSeek V3.2/R1.
 
 Note that the `claude` series has been trained on draw.io diagrams with cloud architecture logos like AWS, Azure, GCP. So if you want to create cloud architecture diagrams, this is the best choice.

app/[lang]/about/cn/page.tsx

@@ -292,6 +292,7 @@ export default function AboutCN() {
                         </li>
                         <li>Anthropic</li>
                         <li>Google AI</li>
+                        <li>Google Vertex AI</li>
                         <li>Azure OpenAI</li>
                         <li>Ollama</li>
                         <li>OpenRouter</li>

app/[lang]/about/ja/page.tsx

@@ -307,6 +307,7 @@ export default function AboutJA() {
                         </li>
                         <li>Anthropic</li>
                         <li>Google AI</li>
+                        <li>Google Vertex AI</li>
                         <li>Azure OpenAI</li>
                         <li>Ollama</li>
                         <li>OpenRouter</li>

app/[lang]/about/page.tsx

@@ -326,6 +326,7 @@ export default function About() {
                         </li>
                         <li>Anthropic</li>
                         <li>Google AI</li>
+                        <li>Google Vertex AI</li>
                         <li>Azure OpenAI</li>
                         <li>Ollama</li>
                         <li>OpenRouter</li>

app/api/chat/route.ts

@@ -34,6 +34,7 @@ import {
     setTraceOutput,
     wrapWithObserve,
 } from "@/lib/langfuse"
+import { findServerModelById } from "@/lib/server-model-config"
 import { getSystemPrompt } from "@/lib/system-prompts"
 import { getUserIdFromRequest } from "@/lib/user-id"
 
@@ -168,6 +169,7 @@ async function handleChatRequest(req: Request): Promise<Response> {
     // Read client AI provider overrides from headers
     const provider = req.headers.get("x-ai-provider")
     let baseUrl = req.headers.get("x-ai-base-url")
+    const selectedModelId = req.headers.get("x-selected-model-id")
 
     // For EdgeOne provider, construct full URL from request origin
     // because createOpenAI needs absolute URL, not relative path
@@ -179,8 +181,30 @@ async function handleChatRequest(req: Request): Promise<Response> {
     // Get cookie header for EdgeOne authentication (eo_token, eo_time)
     const cookieHeader = req.headers.get("cookie")
 
+    // Check if this is a server model with custom env var names
+    let serverModelConfig: {
+        apiKeyEnv?: string
+        baseUrlEnv?: string
+        provider?: string
+    } = {}
+    if (selectedModelId?.startsWith("server:")) {
+        const serverModel = await findServerModelById(selectedModelId)
+        console.log(
+            `[Server Model Lookup] ID: ${selectedModelId}, Found: ${!!serverModel}, Provider: ${serverModel?.provider}`,
+        )
+        if (serverModel) {
+            serverModelConfig = {
+                apiKeyEnv: serverModel.apiKeyEnv,
+                baseUrlEnv: serverModel.baseUrlEnv,
+                // Use actual provider from config (client header may have incorrect value due to ID format change)
+                provider: serverModel.provider,
+            }
+        }
+    }
+
     const clientOverrides = {
-        provider,
+        // Server model provider takes precedence over client header
+        provider: serverModelConfig.provider || provider,
         baseUrl,
         apiKey: req.headers.get("x-ai-api-key"),
         modelId: req.headers.get("x-ai-model"),
@@ -189,6 +213,10 @@ async function handleChatRequest(req: Request): Promise<Response> {
         awsSecretAccessKey: req.headers.get("x-aws-secret-access-key"),
         awsRegion: req.headers.get("x-aws-region"),
         awsSessionToken: req.headers.get("x-aws-session-token"),
+        // Server model custom env var names
+        ...serverModelConfig,
+        // Vertex AI credentials (Express Mode)
+        vertexApiKey: req.headers.get("x-vertex-api-key"),
         // Pass cookies for EdgeOne Pages authentication
         ...(provider === "edgeone" &&
             cookieHeader && {
@@ -199,6 +227,10 @@ async function handleChatRequest(req: Request): Promise<Response> {
     // Read minimal style preference from header
     const minimalStyle = req.headers.get("x-minimal-style") === "true"
 
+    console.log(
+        `[Client Overrides] provider: ${clientOverrides.provider}, modelId: ${clientOverrides.modelId}`,
+    )
+
     // Get AI model with optional client overrides
     const { model, providerOptions, headers, modelId } =
         getAIModel(clientOverrides)
@@ -441,6 +473,13 @@ ${userInputText}
                         inputToRepair = inputToRepair.replace(/:=/g, ": ")
                         // Fix `= "` instead of `: "`
                         inputToRepair = inputToRepair.replace(/=\s*"/g, ': "')
+                        // Fix inconsistent quote escaping in XML attributes within JSON strings
+                        // Pattern: attribute="value\" where opening quote is unescaped but closing is escaped
+                        // Example: y="-20\" should be y=\"-20\"
+                        inputToRepair = inputToRepair.replace(
+                            /(\w+)="([^"]*?)\\"/g,
+                            '$1=\\"$2\\"',
+                        )
                     }
                     // Use jsonrepair to fix truncated JSON
                     const repairedInput = jsonrepair(inputToRepair)

app/api/server-models/route.ts

@@ -0,0 +1,14 @@
+import { NextResponse } from "next/server"
+import { loadFlattenedServerModels } from "@/lib/server-model-config"
+
+// Use dynamic rendering to read AI_MODEL/AI_PROVIDER env vars at runtime
+// This ensures Docker users can set these values when starting containers
+export const dynamic = "force-dynamic"
+
+export async function GET() {
+    const models = await loadFlattenedServerModels()
+    return NextResponse.json({
+        models,
+        hasConfig: models.length > 0,
+    })
+}

app/api/validate-model/route.ts

@@ -3,6 +3,7 @@ import { createAnthropic } from "@ai-sdk/anthropic"
 import { createDeepSeek, deepseek } from "@ai-sdk/deepseek"
 import { createGateway } from "@ai-sdk/gateway"
 import { createGoogleGenerativeAI } from "@ai-sdk/google"
+import { createVertex } from "@ai-sdk/google-vertex"
 import { createOpenAI } from "@ai-sdk/openai"
 import { createOpenRouter } from "@openrouter/ai-sdk-provider"
 import { generateText } from "ai"
@@ -80,6 +81,8 @@ interface ValidateRequest {
     awsAccessKeyId?: string
     awsSecretAccessKey?: string
     awsRegion?: string
+    // Vertex AI specific
+    vertexApiKey?: string // Express Mode API key
 }
 
 export async function POST(req: Request) {
@@ -93,6 +96,8 @@ export async function POST(req: Request) {
             awsAccessKeyId,
             awsSecretAccessKey,
             awsRegion,
+            // Note: Express Mode only needs vertexApiKey
+            vertexApiKey,
         } = body
 
         if (!provider || !modelId) {
@@ -121,6 +126,16 @@ export async function POST(req: Request) {
                     { status: 400 },
                 )
             }
+        } else if (provider === "vertexai") {
+            if (!vertexApiKey) {
+                return NextResponse.json(
+                    {
+                        valid: false,
+                        error: "Vertex AI API key is required for Express Mode",
+                    },
+                    { status: 400 },
+                )
+            }
         } else if (provider !== "ollama" && provider !== "edgeone" && !apiKey) {
             return NextResponse.json(
                 { valid: false, error: "API key is required" },
@@ -158,6 +173,15 @@ export async function POST(req: Request) {
                 break
             }
 
+            case "vertexai": {
+                const vertex = createVertex({
+                    apiKey: vertexApiKey,
+                    ...(baseUrl && { baseURL: baseUrl }),
+                })
+                model = vertex(modelId)
+                break
+            }
+
             case "azure": {
                 const azure = createOpenAI({
                     apiKey,

components/ai-elements/model-selector.tsx

@@ -169,3 +169,27 @@ export const ModelSelectorName = ({
 }: ModelSelectorNameProps) => (
     <span className={cn("flex-1 truncate text-left", className)} {...props} />
 )
+
+export type ModelSelectorSectionHeaderProps = {
+    icon: ReactNode
+    label: string
+    className?: string
+}
+
+export const ModelSelectorSectionHeader = ({
+    icon,
+    label,
+    className,
+}: ModelSelectorSectionHeaderProps) => (
+    <div
+        className={cn(
+            "flex items-center gap-2 px-2 py-1.5 text-xs font-semibold text-muted-foreground bg-muted/40 rounded-sm mx-1 mt-1",
+            className,
+        )}
+    >
+        <span className="[&>svg]:size-3.5" aria-hidden="true">
+            {icon}
+        </span>
+        <span>{label}</span>
+    </div>
+)

components/chat-input.tsx

@@ -27,6 +27,7 @@ import { isPdfFile, isTextFile } from "@/lib/pdf-utils"
 import { STORAGE_KEYS } from "@/lib/storage"
 import type { FlattenedModel } from "@/lib/types/model-config"
 import { extractUrlContent, type UrlData } from "@/lib/url-utils"
+import { isRealDiagram } from "@/lib/utils"
 import { FilePreviewList } from "./file-preview-list"
 
 const MAX_IMAGE_SIZE = 2 * 1024 * 1024 // 2MB
@@ -181,6 +182,7 @@ export function ChatInput({
 }: ChatInputProps) {
     const dict = useDictionary()
     const {
+        chartXML,
         diagramHistory,
         saveDiagramToFile,
         showSaveDialog,
@@ -454,7 +456,7 @@ export function ChatInput({
                             variant="ghost"
                             size="sm"
                             onClick={() => setShowSaveDialog(true)}
-                            disabled={isDisabled}
+                            disabled={isDisabled || !isRealDiagram(chartXML)}
                             tooltipContent={dict.chat.saveDiagram}
                             className="h-8 w-8 p-0 text-muted-foreground hover:text-foreground"
                         >

components/chat-panel.tsx

@@ -963,6 +963,14 @@ export default function ChatPanel({
                         ...(config.awsSessionToken && {
                             "x-aws-session-token": config.awsSessionToken,
                         }),
+                        // Vertex AI credentials (Express Mode)
+                        ...(config.vertexApiKey && {
+                            "x-vertex-api-key": config.vertexApiKey,
+                        }),
+                    }),
+                    // Send selected model ID for server model lookup (apiKeyEnv/baseUrlEnv)
+                    ...(config.selectedModelId && {
+                        "x-selected-model-id": config.selectedModelId,
                     }),
                     ...(minimalStyle && {
                         "x-minimal-style": "true",

components/model-config-dialog.tsx

@@ -78,6 +78,7 @@ const PROVIDER_LOGO_MAP: Record<string, string> = {
     sglang: "openai", // SGLang is OpenAI-compatible
     gateway: "vercel",
     edgeone: "tencent-cloud",
+    vertexai: "google",
     doubao: "bytedance",
     modelscope: "modelscope",
 }
@@ -237,6 +238,7 @@ export function ModelConfigDialog({
             "awsAccessKeyId",
             "awsSecretAccessKey",
             "awsRegion",
+            "vertexApiKey",
         ]
         if (credentialFields.includes(field)) {
             setValidationStatus("idle")
@@ -280,6 +282,7 @@ export function ModelConfigDialog({
         // Check credentials based on provider type
         const isBedrock = selectedProvider.provider === "bedrock"
         const isEdgeOne = selectedProvider.provider === "edgeone"
+        const isVertexAI = selectedProvider.provider === "vertexai"
         if (isBedrock) {
             if (
                 !selectedProvider.awsAccessKeyId ||
@@ -288,6 +291,11 @@ export function ModelConfigDialog({
             ) {
                 return
             }
+        } else if (isVertexAI) {
+            // Vertex AI requires vertexApiKey for Express Mode
+            if (!selectedProvider.vertexApiKey) {
+                return
+            }
         } else if (!isEdgeOne && !selectedProvider.apiKey) {
             return
         }
@@ -328,6 +336,8 @@ export function ModelConfigDialog({
                         awsAccessKeyId: selectedProvider.awsAccessKeyId,
                         awsSecretAccessKey: selectedProvider.awsSecretAccessKey,
                         awsRegion: selectedProvider.awsRegion,
+                        // Vertex AI credentials (Express Mode)
+                        vertexApiKey: selectedProvider.vertexApiKey,
                     }),
                 })
                 const data = await response.json()
@@ -867,7 +877,153 @@ export function ModelConfigDialog({
                                                     </div>
                                                 </>
                                             ) : selectedProvider.provider ===
-                                              "edgeone" ? (
+                                              "vertexai" ? (
+                                                <>
+                                                    {/* Vertex AI API Key */}
+                                                    <div className="space-y-2">
+                                                        <Label
+                                                            htmlFor="vertex-api-key"
+                                                            className="text-xs font-medium flex items-center gap-1.5"
+                                                        >
+                                                            <Key className="h-3.5 w-3.5 text-muted-foreground" />
+                                                            API Key
+                                                        </Label>
+                                                        <div className="flex gap-2">
+                                                            <div className="relative flex-1">
+                                                                <Input
+                                                                    id="vertex-api-key"
+                                                                    type={
+                                                                        showApiKey
+                                                                            ? "text"
+                                                                            : "password"
+                                                                    }
+                                                                    value={
+                                                                        selectedProvider.vertexApiKey ||
+                                                                        ""
+                                                                    }
+                                                                    onChange={(
+                                                                        e,
+                                                                    ) =>
+                                                                        handleProviderUpdate(
+                                                                            "vertexApiKey",
+                                                                            e
+                                                                                .target
+                                                                                .value,
+                                                                        )
+                                                                    }
+                                                                    placeholder="Enter your Vertex AI API key"
+                                                                    className="h-9 pr-10 font-mono text-xs"
+                                                                />
+                                                                <button
+                                                                    type="button"
+                                                                    onClick={() =>
+                                                                        setShowApiKey(
+                                                                            !showApiKey,
+                                                                        )
+                                                                    }
+                                                                    aria-label={
+                                                                        showApiKey
+                                                                            ? "Hide API key"
+                                                                            : "Show API key"
+                                                                    }
+                                                                    className="absolute right-3 top-1/2 -translate-y-1/2 text-muted-foreground hover:text-foreground transition-colors focus-visible:outline-none focus-visible:ring-2 focus-visible:ring-ring focus-visible:ring-offset-2 rounded"
+                                                                >
+                                                                    {showApiKey ? (
+                                                                        <EyeOff className="h-4 w-4" />
+                                                                    ) : (
+                                                                        <Eye className="h-4 w-4" />
+                                                                    )}
+                                                                </button>
+                                                            </div>
+                                                            <Button
+                                                                variant={
+                                                                    validationStatus ===
+                                                                    "success"
+                                                                        ? "outline"
+                                                                        : "default"
+                                                                }
+                                                                size="sm"
+                                                                onClick={
+                                                                    handleValidate
+                                                                }
+                                                                disabled={
+                                                                    !selectedProvider.vertexApiKey ||
+                                                                    validationStatus ===
+                                                                        "validating"
+                                                                }
+                                                                className={cn(
+                                                                    "h-9 px-4",
+                                                                    validationStatus ===
+                                                                        "success" &&
+                                                                        "text-success border-success/30 bg-success-muted hover:bg-success-muted",
+                                                                )}
+                                                            >
+                                                                {validationStatus ===
+                                                                "validating" ? (
+                                                                    <Loader2 className="h-4 w-4 animate-spin" />
+                                                                ) : validationStatus ===
+                                                                  "success" ? (
+                                                                    <>
+                                                                        <Check className="h-4 w-4 mr-1.5 animate-check-pop" />
+                                                                        {
+                                                                            dict
+                                                                                .modelConfig
+                                                                                .verified
+                                                                        }
+                                                                    </>
+                                                                ) : (
+                                                                    dict
+                                                                        .modelConfig
+                                                                        .test
+                                                                )}
+                                                            </Button>
+                                                        </div>
+                                                        {validationStatus ===
+                                                            "error" &&
+                                                            validationError && (
+                                                                <p className="text-xs text-destructive flex items-center gap-1">
+                                                                    <X className="h-3 w-3" />
+                                                                    {
+                                                                        validationError
+                                                                    }
+                                                                </p>
+                                                            )}
+                                                    </div>
+
+                                                    {/* Base URL (optional) */}
+                                                    <div className="space-y-2">
+                                                        <Label
+                                                            htmlFor="vertex-base-url"
+                                                            className="text-xs font-medium flex items-center gap-1.5"
+                                                        >
+                                                            <Link2 className="h-3.5 w-3.5 text-muted-foreground" />
+                                                            Base URL{" "}
+                                                            <span className="text-muted-foreground font-normal">
+                                                                (optional)
+                                                            </span>
+                                                        </Label>
+                                                        <Input
+                                                            id="vertex-base-url"
+                                                            value={
+                                                                selectedProvider.baseUrl ||
+                                                                ""
+                                                            }
+                                                            onChange={(e) =>
+                                                                handleProviderUpdate(
+                                                                    "baseUrl",
+                                                                    e.target
+                                                                        .value,
+                                                                )
+                                                            }
+                                                            placeholder="Custom endpoint URL"
+                                                            className="h-9 font-mono text-xs"
+                                                        />
+                                                    </div>
+                                                </>
+                                            ) : selectedProvider.provider ===
+                                                  "ollama" ||
+                                              selectedProvider.provider ===
+                                                  "edgeone" ? (
                                                 <div className="space-y-3">
                                                     <div className="flex items-center gap-2">
                                                         <Button

components/model-selector.tsx

@@ -5,8 +5,10 @@ import {
     Bot,
     Check,
     ChevronDown,
+    Monitor,
     Server,
     Settings2,
+    User,
 } from "lucide-react"
 import { useEffect, useMemo, useRef, useState } from "react"
 import {
@@ -19,6 +21,7 @@ import {
     ModelSelectorLogo,
     ModelSelectorName,
     ModelSelector as ModelSelectorRoot,
+    ModelSelectorSectionHeader,
     ModelSelectorSeparator,
     ModelSelectorTrigger,
 } from "@/components/ai-elements/model-selector"
@@ -49,6 +52,7 @@ const PROVIDER_LOGO_MAP: Record<string, string> = {
     sglang: "openai", // SGLang is OpenAI-compatible, use OpenAI logo
     gateway: "vercel",
     edgeone: "tencent-cloud",
+    vertexai: "google",
     doubao: "bytedance",
     modelscope: "modelscope",
 }
@@ -62,7 +66,11 @@ function groupModelsByProvider(
         { provider: string; models: FlattenedModel[] }
     >()
     for (const model of models) {
-        const key = model.providerLabel
+        // For server models, strip "Server · " prefix for cleaner grouping
+        const key =
+            model.source === "server"
+                ? model.providerLabel.replace(/^Server · /, "")
+                : model.providerLabel
         const existing = groups.get(key)
         if (existing) {
             existing.models.push(model)
@@ -90,10 +98,26 @@ export function ModelSelector({
         }
         return models.filter((m) => m.validated === true)
     }, [models, showUnvalidatedModels])
-    const groupedModels = useMemo(
-        () => groupModelsByProvider(displayModels),
+
+    // Separate server and user models
+    const serverModels = useMemo(
+        () => displayModels.filter((m) => m.source === "server"),
         [displayModels],
     )
+    const userModels = useMemo(
+        () => displayModels.filter((m) => m.source !== "server"),
+        [displayModels],
+    )
+
+    // Group each category separately
+    const groupedServerModels = useMemo(
+        () => groupModelsByProvider(serverModels),
+        [serverModels],
+    )
+    const groupedUserModels = useMemo(
+        () => groupModelsByProvider(userModels),
+        [userModels],
+    )
 
     // Find selected model for display
     const selectedModel = useMemo(
@@ -160,7 +184,7 @@ export function ModelSelector({
                         size="sm"
                         disabled={disabled}
                         className={cn(
-                            "hover:bg-accent gap-1.5 h-8 px-2 transition-all duration-150 ease-in-out",
+                            "hover:bg-accent gap-1.5 h-8 px-2 transition-[padding,background-color] duration-150 ease-in-out",
                             !showLabel && "px-1.5 justify-center",
                         )}
                         // accessibility: expose label to screen readers
@@ -197,83 +221,169 @@ export function ModelSelector({
                                 : dict.modelConfig.noModelsFound}
                         </ModelSelectorEmpty>
 
-                        {/* Server Default Option */}
-                        <ModelSelectorGroup heading={dict.modelConfig.default}>
-                            <ModelSelectorItem
-                                value="__server_default__"
-                                onSelect={handleSelect}
-                                className={cn(
-                                    "cursor-pointer",
-                                    !selectedModelId && "bg-accent",
-                                )}
+                        {/* Server Default Option - only show when no server models are configured */}
+                        {serverModels.length === 0 && (
+                            <ModelSelectorGroup
+                                heading={dict.modelConfig.default}
                             >
-                                <Check
+                                <ModelSelectorItem
+                                    value="__server_default__"
+                                    onSelect={handleSelect}
                                     className={cn(
-                                        "mr-2 h-4 w-4",
-                                        !selectedModelId
-                                            ? "opacity-100"
-                                            : "opacity-0",
+                                        "cursor-pointer",
+                                        !selectedModelId && "bg-accent",
                                     )}
-                                />
-                                <Server className="mr-2 h-4 w-4 text-muted-foreground" />
-                                <ModelSelectorName>
-                                    {dict.modelConfig.serverDefault}
-                                </ModelSelectorName>
-                            </ModelSelectorItem>
-                        </ModelSelectorGroup>
-
-                        {/* Configured Models by Provider */}
-                        {Array.from(groupedModels.entries()).map(
-                            ([
-                                providerLabel,
-                                { provider, models: providerModels },
-                            ]) => (
-                                <ModelSelectorGroup
-                                    key={providerLabel}
-                                    heading={providerLabel}
                                 >
-                                    {providerModels.map((model) => (
-                                        <ModelSelectorItem
-                                            key={model.id}
-                                            value={model.modelId}
-                                            onSelect={() =>
-                                                handleSelect(model.id)
-                                            }
-                                            className="cursor-pointer"
+                                    <Check
+                                        className={cn(
+                                            "mr-2 h-4 w-4",
+                                            !selectedModelId
+                                                ? "opacity-100"
+                                                : "opacity-0",
+                                        )}
+                                    />
+                                    <Server className="mr-2 h-4 w-4 text-muted-foreground" />
+                                    <ModelSelectorName>
+                                        {dict.modelConfig.serverDefault}
+                                    </ModelSelectorName>
+                                </ModelSelectorItem>
+                            </ModelSelectorGroup>
+                        )}
+
+                        {/* Server Models Section */}
+                        {serverModels.length > 0 && (
+                            <>
+                                <ModelSelectorSectionHeader
+                                    icon={<Monitor />}
+                                    label={dict.modelConfig.serverModels}
+                                />
+                                {Array.from(groupedServerModels.entries()).map(
+                                    ([
+                                        providerLabel,
+                                        { provider, models: providerModels },
+                                    ]) => (
+                                        <ModelSelectorGroup
+                                            key={`server-${providerLabel}`}
+                                            heading={providerLabel}
+                                            className="[&>[cmdk-group-heading]]:pl-4"
                                         >
-                                            <Check
-                                                className={cn(
-                                                    "mr-2 h-4 w-4",
-                                                    selectedModelId === model.id
-                                                        ? "opacity-100"
-                                                        : "opacity-0",
-                                                )}
-                                            />
-                                            <ModelSelectorLogo
-                                                provider={
-                                                    PROVIDER_LOGO_MAP[
-                                                        provider
-                                                    ] || provider
-                                                }
-                                                className="mr-2"
-                                            />
-                                            <ModelSelectorName>
-                                                {model.modelId}
-                                            </ModelSelectorName>
-                                            {model.validated !== true && (
-                                                <span
-                                                    title={
-                                                        dict.modelConfig
-                                                            .unvalidatedModelWarning
+                                            {providerModels.map((model) => (
+                                                <ModelSelectorItem
+                                                    key={model.id}
+                                                    value={model.modelId}
+                                                    onSelect={() =>
+                                                        handleSelect(model.id)
                                                     }
+                                                    className="cursor-pointer"
                                                 >
-                                                    <AlertTriangle className="ml-auto h-3 w-3 text-warning" />
-                                                </span>
-                                            )}
-                                        </ModelSelectorItem>
-                                    ))}
-                                </ModelSelectorGroup>
-                            ),
+                                                    <Check
+                                                        className={cn(
+                                                            "mr-2 h-4 w-4",
+                                                            selectedModelId ===
+                                                                model.id
+                                                                ? "opacity-100"
+                                                                : "opacity-0",
+                                                        )}
+                                                    />
+                                                    <ModelSelectorLogo
+                                                        provider={
+                                                            PROVIDER_LOGO_MAP[
+                                                                provider
+                                                            ] || provider
+                                                        }
+                                                        className="mr-2"
+                                                    />
+                                                    <ModelSelectorName>
+                                                        {model.modelId}
+                                                    </ModelSelectorName>
+                                                    {model.isDefault && (
+                                                        <span
+                                                            title={
+                                                                dict.modelConfig
+                                                                    .serverDefaultModel
+                                                            }
+                                                            className="ml-auto text-xs text-muted-foreground"
+                                                        >
+                                                            {
+                                                                dict.modelConfig
+                                                                    .default
+                                                            }
+                                                        </span>
+                                                    )}
+                                                </ModelSelectorItem>
+                                            ))}
+                                        </ModelSelectorGroup>
+                                    ),
+                                )}
+                            </>
+                        )}
+
+                        {/* User Models Section */}
+                        {userModels.length > 0 && (
+                            <>
+                                {serverModels.length > 0 && (
+                                    <ModelSelectorSeparator />
+                                )}
+                                <ModelSelectorSectionHeader
+                                    icon={<User />}
+                                    label={dict.modelConfig.userModels}
+                                />
+                                {Array.from(groupedUserModels.entries()).map(
+                                    ([
+                                        providerLabel,
+                                        { provider, models: providerModels },
+                                    ]) => (
+                                        <ModelSelectorGroup
+                                            key={`user-${providerLabel}`}
+                                            heading={providerLabel}
+                                            className="[&>[cmdk-group-heading]]:pl-4"
+                                        >
+                                            {providerModels.map((model) => (
+                                                <ModelSelectorItem
+                                                    key={model.id}
+                                                    value={model.modelId}
+                                                    onSelect={() =>
+                                                        handleSelect(model.id)
+                                                    }
+                                                    className="cursor-pointer"
+                                                >
+                                                    <Check
+                                                        className={cn(
+                                                            "mr-2 h-4 w-4",
+                                                            selectedModelId ===
+                                                                model.id
+                                                                ? "opacity-100"
+                                                                : "opacity-0",
+                                                        )}
+                                                    />
+                                                    <ModelSelectorLogo
+                                                        provider={
+                                                            PROVIDER_LOGO_MAP[
+                                                                provider
+                                                            ] || provider
+                                                        }
+                                                        className="mr-2"
+                                                    />
+                                                    <ModelSelectorName>
+                                                        {model.modelId}
+                                                    </ModelSelectorName>
+                                                    {model.validated !==
+                                                        true && (
+                                                        <span
+                                                            title={
+                                                                dict.modelConfig
+                                                                    .unvalidatedModelWarning
+                                                            }
+                                                        >
+                                                            <AlertTriangle className="ml-auto h-3 w-3 text-warning" />
+                                                        </span>
+                                                    )}
+                                                </ModelSelectorItem>
+                                            ))}
+                                        </ModelSelectorGroup>
+                                    ),
+                                )}
+                            </>
                         )}
 
                         {/* Configure Option */}
@@ -282,7 +392,7 @@ export function ModelSelector({
                             <ModelSelectorItem
                                 value="__configure__"
                                 onSelect={handleSelect}
-                                className="cursor-pointer"
+                                className="cursor-pointer text-muted-foreground hover:text-foreground"
                             >
                                 <Settings2 className="mr-2 h-4 w-4" />
                                 <ModelSelectorName>

docs/cn/README_CN.md

@@ -200,6 +200,7 @@ npm run dev
 -   OpenAI
 -   Anthropic
 -   Google AI
+-   Google Vertex AI
 -   Azure OpenAI
 -   Ollama
 -   OpenRouter
@@ -213,6 +214,10 @@ npm run dev
 
 📖 **[详细的提供商配置指南](./ai-providers.md)** - 查看各提供商的设置说明。
 
+### 服务端多模型配置
+
+管理员可以配置多个服务端模型，让所有用户无需提供个人 API Key 即可使用。通过 `AI_MODELS_CONFIG` 环境变量（JSON 字符串）或 `ai-models.json` 文件配置。
+
 **模型要求**：此任务需要强大的模型能力，因为它涉及生成具有严格格式约束的长文本（draw.io XML）。推荐使用 Claude Sonnet 4.5、GPT-5.1、Gemini 3 Pro 和 DeepSeek V3.2/R1。
 
 注意：`claude` 系列已在带有 AWS、Azure、GCP 等云架构 Logo 的 draw.io 图表上进行训练，因此如果您想创建云架构图，这是最佳选择。

docs/cn/ai-providers.md

@@ -217,6 +217,63 @@ AI_MODEL=openai/gpt-4o
 AI_PROVIDER=google  # 或：openai, anthropic, deepseek, siliconflow, doubao, azure, bedrock, openrouter, ollama, gateway, sglang
 ```
 
+## 服务端多模型配置
+
+管理员可以配置多个服务端模型，让所有用户无需提供个人 API Key 即可使用。
+
+### 配置方式
+
+**方式一：环境变量**（推荐用于云部署）
+
+设置 `AI_MODELS_CONFIG` 为 JSON 字符串：
+
+```bash
+AI_MODELS_CONFIG='{"providers":[{"name":"OpenAI","provider":"openai","models":["gpt-4o"],"default":true}]}'
+```
+
+**方式二：配置文件**
+
+在项目根目录创建 `ai-models.json` 文件（或通过 `AI_MODELS_CONFIG_PATH` 指定路径）。
+
+### 配置示例
+
+```json
+{
+  "providers": [
+    {
+      "name": "OpenAI Production",
+      "provider": "openai",
+      "models": ["gpt-4o", "gpt-4o-mini"],
+      "default": true
+    },
+    {
+      "name": "Custom DeepSeek",
+      "provider": "deepseek",
+      "models": ["deepseek-chat"],
+      "apiKeyEnv": "MY_DEEPSEEK_KEY",
+      "baseUrlEnv": "MY_DEEPSEEK_URL"
+    }
+  ]
+}
+```
+
+### 字段说明
+
+| 字段 | 必填 | 说明 |
+|------|------|------|
+| `name` | 是 | 显示名称（支持同一提供商多个配置） |
+| `provider` | 是 | 提供商类型（`openai`, `anthropic`, `google`, `bedrock` 等） |
+| `models` | 是 | 模型 ID 列表 |
+| `default` | 否 | 设为 `true` 表示默认选中该提供商的第一个模型 |
+| `apiKeyEnv` | 否 | 自定义 API Key 环境变量名（默认使用提供商标准变量如 `OPENAI_API_KEY`） |
+| `baseUrlEnv` | 否 | 自定义 Base URL 环境变量名 |
+
+### 说明
+
+- API Key 和凭证通过环境变量提供。默认使用标准变量名（如 `OPENAI_API_KEY`），也可通过 `apiKeyEnv` 指定自定义变量名。
+- `name` 字段允许同一提供商多个配置（例如 "OpenAI Production" 和 "OpenAI Staging" 都使用 `provider: "openai"` 但 `apiKeyEnv` 不同）。
+- 如果配置不存在，应用会回退到 `AI_PROVIDER`/`AI_MODEL` 环境变量配置。
+
 ## 模型能力要求
 
 此任务对模型能力要求极高，因为它涉及生成具有严格格式约束（draw.io XML）的长文本。

docs/en/ai-providers.md

@@ -33,6 +33,21 @@ Optional custom endpoint:
 GOOGLE_BASE_URL=https://your-custom-endpoint
 ```
 
+### Google Vertex AI (Enterprise GCP)
+
+Google Vertex AI offers enterprise-grade features and data residency. **Express Mode** allows for simple API key authentication, making it compatible with edge runtimes like Vercel and Cloudflare.
+
+```bash
+GOOGLE_VERTEX_API_KEY=your_api_key
+AI_MODEL=gemini-2.0-flash
+```
+
+Optional custom endpoint:
+
+```bash
+GOOGLE_VERTEX_BASE_URL=https://your-custom-endpoint
+```
+
 ### OpenAI
 
 ```bash
@@ -217,6 +232,63 @@ If you configure **multiple** API keys, you must explicitly set `AI_PROVIDER`:
 AI_PROVIDER=google  # or: openai, anthropic, deepseek, siliconflow, doubao, azure, bedrock, openrouter, ollama, gateway, sglang, modelscope
 ```
 
+## Server-Side Multi-Model Configuration
+
+Administrators can configure multiple server-side models that are available to all users without requiring personal API keys.
+
+### Configuration Methods
+
+**Option 1: Environment Variable** (recommended for cloud deployments)
+
+Set `AI_MODELS_CONFIG` as a JSON string:
+
+```bash
+AI_MODELS_CONFIG='{"providers":[{"name":"OpenAI","provider":"openai","models":["gpt-4o"],"default":true}]}'
+```
+
+**Option 2: Config File**
+
+Create an `ai-models.json` file in the project root (or set `AI_MODELS_CONFIG_PATH` to a custom location).
+
+### Example Configuration
+
+```json
+{
+  "providers": [
+    {
+      "name": "OpenAI Production",
+      "provider": "openai",
+      "models": ["gpt-4o", "gpt-4o-mini"],
+      "default": true
+    },
+    {
+      "name": "Custom DeepSeek",
+      "provider": "deepseek",
+      "models": ["deepseek-chat"],
+      "apiKeyEnv": "MY_DEEPSEEK_KEY",
+      "baseUrlEnv": "MY_DEEPSEEK_URL"
+    }
+  ]
+}
+```
+
+### Field Reference
+
+| Field | Required | Description |
+|-------|----------|-------------|
+| `name` | Yes | Display name (supports multiple configs for same provider) |
+| `provider` | Yes | Provider type (`openai`, `anthropic`, `google`, `bedrock`, etc.) |
+| `models` | Yes | List of model IDs |
+| `default` | No | Set to `true` to auto-select this provider's first model as default |
+| `apiKeyEnv` | No | Custom API key env var name (defaults to provider's standard var like `OPENAI_API_KEY`) |
+| `baseUrlEnv` | No | Custom base URL env var name |
+
+### Notes
+
+- API keys and credentials are provided via environment variables. By default, standard var names are used (e.g., `OPENAI_API_KEY`), but you can specify custom var names with `apiKeyEnv`.
+- The `name` field allows multiple configurations for the same provider (e.g., "OpenAI Production" and "OpenAI Staging" both using `provider: "openai"` but with different `apiKeyEnv` values).
+- If config is not present, the app falls back to `AI_PROVIDER`/`AI_MODEL` environment variable configuration.
+
 ## Model Capability Requirements
 
 This task requires exceptionally strong model capabilities, as it involves generating long-form text with strict formatting constraints (draw.io XML).

docs/en/docker.md

@@ -22,6 +22,27 @@ cp env.example .env
 docker run -d -p 3000:3000 --env-file .env ghcr.io/dayuanjiang/next-ai-draw-io:latest
 ```
 
+### Using server-side model configuration
+
+You can mount an `ai-models.json` file into the container to provide multiple server-side models without exposing user API keys:
+
+```bash
+docker run -d -p 3000:3000 \
+  -e OPENAI_API_KEY=your_api_key \
+  -v $(pwd)/ai-models.json:/app/ai-models.json:ro \
+  ghcr.io/dayuanjiang/next-ai-draw-io:latest
+```
+
+If you prefer to keep the config in a different path inside the container, set `AI_MODELS_CONFIG_PATH`:
+
+```bash
+docker run -d -p 3000:3000 \
+  -e OPENAI_API_KEY=your_api_key \
+  -e AI_MODELS_CONFIG_PATH=/config/ai-models.json \
+  -v $(pwd)/ai-models.json:/config/ai-models.json:ro \
+  ghcr.io/dayuanjiang/next-ai-draw-io:latest
+```
+
 Open [http://localhost:3000](http://localhost:3000) in your browser.
 
 Replace the environment variables with your preferred AI provider configuration. See [AI Providers](./ai-providers.md) for available options.

docs/ja/README_JA.md

@@ -201,6 +201,7 @@ Next.jsアプリをデプロイする最も簡単な方法は、Next.jsの作成
 -   OpenAI
 -   Anthropic
 -   Google AI
+-   Google Vertex AI
 -   Azure OpenAI
 -   Ollama
 -   OpenRouter
@@ -214,6 +215,10 @@ AWS BedrockとOpenRouter以外のすべてのプロバイダーはカスタム
 
 📖 **[詳細なプロバイダー設定ガイド](./ai-providers.md)** - 各プロバイダーの設定手順をご覧ください。
 
+### サーバーサイドマルチモデル設定
+
+管理者は、ユーザーが個人のAPIキーを提供することなく利用できる複数のサーバーサイドモデルを設定できます。`AI_MODELS_CONFIG` 環境変数（JSON文字列）または `ai-models.json` ファイルで設定します。
+
 **モデル要件**：このタスクは厳密なフォーマット制約（draw.io XML）を持つ長文テキスト生成を伴うため、強力なモデル機能が必要です。Claude Sonnet 4.5、GPT-5.1、Gemini 3 Pro、DeepSeek V3.2/R1を推奨します。
 
 注：`claude`シリーズはAWS、Azure、GCPなどのクラウドアーキテクチャロゴ付きのdraw.ioダイアグラムで学習されているため、クラウドアーキテクチャダイアグラムを作成したい場合は最適な選択です。

docs/ja/ai-providers.md

@@ -217,6 +217,63 @@ AI_MODEL=openai/gpt-4o
 AI_PROVIDER=google  # または: openai, anthropic, deepseek, siliconflow, doubao, azure, bedrock, openrouter, ollama, gateway, sglang
 ```
 
+## サーバーサイドマルチモデル設定
+
+管理者は、ユーザーが個人のAPIキーを提供することなく利用できる複数のサーバーサイドモデルを設定できます。
+
+### 設定方法
+
+**方法1：環境変数**（クラウドデプロイ推奨）
+
+`AI_MODELS_CONFIG` をJSON文字列として設定：
+
+```bash
+AI_MODELS_CONFIG='{"providers":[{"name":"OpenAI","provider":"openai","models":["gpt-4o"],"default":true}]}'
+```
+
+**方法2：設定ファイル**
+
+プロジェクトルートに `ai-models.json` ファイルを作成します（または `AI_MODELS_CONFIG_PATH` でパスを指定）。
+
+### 設定例
+
+```json
+{
+  "providers": [
+    {
+      "name": "OpenAI Production",
+      "provider": "openai",
+      "models": ["gpt-4o", "gpt-4o-mini"],
+      "default": true
+    },
+    {
+      "name": "Custom DeepSeek",
+      "provider": "deepseek",
+      "models": ["deepseek-chat"],
+      "apiKeyEnv": "MY_DEEPSEEK_KEY",
+      "baseUrlEnv": "MY_DEEPSEEK_URL"
+    }
+  ]
+}
+```
+
+### フィールド説明
+
+| フィールド | 必須 | 説明 |
+|------------|------|------|
+| `name` | はい | 表示名（同一プロバイダーの複数設定をサポート） |
+| `provider` | はい | プロバイダータイプ（`openai`, `anthropic`, `google`, `bedrock` など） |
+| `models` | はい | モデルIDのリスト |
+| `default` | いいえ | `true` に設定すると、そのプロバイダーの最初のモデルがデフォルトで選択されます |
+| `apiKeyEnv` | いいえ | カスタムAPIキー環境変数名（デフォルトは `OPENAI_API_KEY` などの標準変数） |
+| `baseUrlEnv` | いいえ | カスタムBase URL環境変数名 |
+
+### 備考
+
+- APIキーと認証情報は環境変数で提供します。デフォルトは標準変数名（例：`OPENAI_API_KEY`）を使用しますが、`apiKeyEnv` でカスタム変数名を指定できます。
+- `name` フィールドにより同一プロバイダーの複数設定が可能です（例：「OpenAI Production」と「OpenAI Staging」が両方とも `provider: "openai"` を使用しつつ、異なる `apiKeyEnv` を持つ）。
+- 設定が存在しない場合、アプリは `AI_PROVIDER`/`AI_MODEL` 環境変数設定にフォールバックします。
+
 ## モデル性能要件
 
 このタスクは、厳密なフォーマット制約（draw.io XML）を伴う長文テキストの生成を含むため、非常に強力なモデル性能が必要です。

env.example

@@ -1,6 +1,6 @@
 # AI Provider Configuration
 # AI_PROVIDER: Which provider to use
-# Options: bedrock, openai, anthropic, google, azure, ollama, openrouter, deepseek, siliconflow, gateway
+# Options: bedrock, openai, anthropic, google, vertexai, azure, ollama, openrouter, deepseek, siliconflow, gateway
 # Default: bedrock
 AI_PROVIDER=bedrock
 
@@ -40,6 +40,14 @@ AI_MODEL=global.anthropic.claude-sonnet-4-5-20250929-v1:0
 # GOOGLE_THINKING_BUDGET=8192                # Optional: Gemini 2.5 thinking budget in tokens (for more/less thinking)
 # GOOGLE_THINKING_LEVEL=high                 # Optional: Gemini 3 thinking level (low/high)
 
+# Google Vertex AI Configuration (Enterprise GCP)
+# For enterprise users needing data residency, VPC Service Controls, or GCP integration
+# GOOGLE_VERTEX_API_KEY=                          # Required: Express Mode API key
+# GOOGLE_VERTEX_BASE_URL=https://...              # Optional: Custom endpoint URL
+# Note: Gemini 2.5/3 models automatically enable reasoning display (includeThoughts: true)
+# GOOGLE_VERTEX_THINKING_BUDGET=8192              # Optional: Gemini 2.5 thinking budget in tokens (1024-100000)
+# GOOGLE_VERTEX_THINKING_LEVEL=high               # Optional: Gemini 3 thinking level (minimal/low/medium/high)
+
 # Azure OpenAI Configuration
 # Configure endpoint using ONE of these methods:
 #   1. AZURE_RESOURCE_NAME - SDK constructs: https://{name}.openai.azure.com/openai/v1{path}
@@ -93,6 +101,11 @@ AI_MODEL=global.anthropic.claude-sonnet-4-5-20250929-v1:0
 # LANGFUSE_SECRET_KEY=sk-lf-...
 # LANGFUSE_BASEURL=https://cloud.langfuse.com  # EU region, use https://us.cloud.langfuse.com for US
 
+# Optional server-side multi-model configuration
+# If set, points to a JSON file with server-provided models (see README for schema).
+# Default: ./ai-models.json in project root
+# AI_MODELS_CONFIG_PATH=/path/to/ai-models.json
+
 # Temperature (Optional)
 # Controls randomness in AI responses. Lower = more deterministic.
 # Leave unset for models that don't support temperature (e.g., GPT-5.1 reasoning models)

hooks/use-model-config.ts

@@ -1,6 +1,7 @@
 "use client"
 
 import { useCallback, useEffect, useState } from "react"
+import type { FlattenedServerModel } from "@/lib/server-model-config"
 import { STORAGE_KEYS } from "@/lib/storage"
 import {
     createEmptyConfig,
@@ -132,14 +133,56 @@ export interface UseModelConfigReturn {
 export function useModelConfig(): UseModelConfigReturn {
     const [config, setConfig] = useState<MultiModelConfig>(createEmptyConfig)
     const [isLoaded, setIsLoaded] = useState(false)
+    const [serverModels, setServerModels] = useState<FlattenedServerModel[]>([])
+    const [serverLoaded, setServerLoaded] = useState(false)
 
-    // Load config on mount
+    // Load client config on mount
     useEffect(() => {
         const loaded = loadConfig()
         setConfig(loaded)
         setIsLoaded(true)
     }, [])
 
+    // Load server models on mount (if any)
+    useEffect(() => {
+        if (typeof window === "undefined") return
+
+        fetch("/api/server-models")
+            .then((res) => {
+                if (!res.ok) {
+                    console.error(
+                        "Failed to load server models:",
+                        res.status,
+                        res.statusText,
+                    )
+                    throw new Error(`Request failed with status ${res.status}`)
+                }
+                return res.json()
+            })
+            .then((data) => {
+                const raw: FlattenedServerModel[] = data?.models || []
+                setServerModels(raw)
+                setServerLoaded(true)
+
+                // Auto-select default server model if no model is currently selected
+                setConfig((prev) => {
+                    if (!prev.selectedModelId && raw.length > 0) {
+                        const defaultModel = raw.find((m) => m.isDefault)
+                        if (defaultModel) {
+                            return { ...prev, selectedModelId: defaultModel.id }
+                        }
+                        // If no default marked, use first server model
+                        return { ...prev, selectedModelId: raw[0].id }
+                    }
+                    return prev
+                })
+            })
+            .catch((error) => {
+                console.error("Error while loading server models:", error)
+                setServerLoaded(true)
+            })
+    }, [])
+
     // Save config whenever it changes (after initial load)
     useEffect(() => {
         if (isLoaded) {
@@ -148,9 +191,33 @@ export function useModelConfig(): UseModelConfigReturn {
     }, [config, isLoaded])
 
     // Derived state
-    const models = flattenModels(config)
+    const userModels = flattenModels(config)
+
+    const models: FlattenedModel[] = [
+        // Server models (read-only, credentials from env)
+        ...serverModels.map((m) => ({
+            id: m.id,
+            modelId: m.modelId,
+            provider: m.provider,
+            providerLabel: `Server · ${m.providerLabel}`,
+            apiKey: "",
+            baseUrl: undefined,
+            awsAccessKeyId: undefined,
+            awsSecretAccessKey: undefined,
+            awsRegion: undefined,
+            awsSessionToken: undefined,
+            validated: true,
+            source: "server" as const,
+            isDefault: m.isDefault,
+            apiKeyEnv: m.apiKeyEnv,
+            baseUrlEnv: m.baseUrlEnv,
+        })),
+        // User models from local configuration
+        ...userModels,
+    ]
+
     const selectedModel = config.selectedModelId
-        ? findModelById(config, config.selectedModelId)
+        ? models.find((m) => m.id === config.selectedModelId)
         : undefined
 
     // Actions
@@ -282,7 +349,7 @@ export function useModelConfig(): UseModelConfigReturn {
 
     return {
         config,
-        isLoaded,
+        isLoaded: isLoaded && serverLoaded,
         models,
         selectedModel,
         selectedModelId: config.selectedModelId,
@@ -314,6 +381,10 @@ export function getSelectedAIConfig(): {
     awsSecretAccessKey: string
     awsRegion: string
     awsSessionToken: string
+    // Selected model ID (for server model lookup)
+    selectedModelId: string
+    // Vertex AI credentials (Express Mode)
+    vertexApiKey: string
 } {
     const empty = {
         accessCode: "",
@@ -325,6 +396,8 @@ export function getSelectedAIConfig(): {
         awsSecretAccessKey: "",
         awsRegion: "",
         awsSessionToken: "",
+        selectedModelId: "",
+        vertexApiKey: "",
     }
 
     if (typeof window === "undefined") return empty
@@ -347,6 +420,8 @@ export function getSelectedAIConfig(): {
             awsSecretAccessKey: "",
             awsRegion: "",
             awsSessionToken: "",
+            selectedModelId: "",
+            vertexApiKey: "",
         }
     }
 
@@ -357,12 +432,32 @@ export function getSelectedAIConfig(): {
         return { ...empty, accessCode }
     }
 
-    // No selected model = use server default
+    // No selected model = use server default (AI_PROVIDER/AI_MODEL/env auto-detect)
     if (!config.selectedModelId) {
         return { ...empty, accessCode }
     }
 
-    // Find selected model
+    // Server-side model selection (id = "server:<name-slug>:<modelId>")
+    // Provider is resolved server-side via findServerModelById()
+    if (config.selectedModelId.startsWith("server:")) {
+        const parts = config.selectedModelId.split(":")
+        const nameSlug = parts[1] || ""
+        const modelId = parts.slice(2).join(":") // Preserve Bedrock-style IDs
+
+        return {
+            ...empty,
+            accessCode,
+            // Note: nameSlug is NOT the provider, but we send it for backwards compat
+            // Server uses selectedModelId to lookup the actual provider
+            aiProvider: nameSlug,
+            aiBaseUrl: "",
+            aiApiKey: "",
+            aiModel: modelId,
+            selectedModelId: config.selectedModelId,
+        }
+    }
+
+    // Find selected user-defined model
     const model = findModelById(config, config.selectedModelId)
     if (!model) {
         return { ...empty, accessCode }
@@ -379,5 +474,8 @@ export function getSelectedAIConfig(): {
         awsSecretAccessKey: model.awsSecretAccessKey || "",
         awsRegion: model.awsRegion || "",
         awsSessionToken: model.awsSessionToken || "",
+        selectedModelId: config.selectedModelId || "",
+        // Vertex AI credentials (Express Mode)
+        vertexApiKey: model.vertexApiKey || "",
     }
 }

lib/ai-providers.ts

@@ -4,6 +4,7 @@ import { azure, createAzure } from "@ai-sdk/azure"
 import { createDeepSeek, deepseek } from "@ai-sdk/deepseek"
 import { createGateway, gateway } from "@ai-sdk/gateway"
 import { createGoogleGenerativeAI, google } from "@ai-sdk/google"
+import { createVertex } from "@ai-sdk/google-vertex"
 import { createOpenAI, openai } from "@ai-sdk/openai"
 import { fromNodeProviderChain } from "@aws-sdk/credential-providers"
 import { createOpenRouter } from "@openrouter/ai-sdk-provider"
@@ -29,8 +30,13 @@ export interface ClientOverrides {
     awsSecretAccessKey?: string | null
     awsRegion?: string | null
     awsSessionToken?: string | null
+    // Vertex AI config
+    vertexApiKey?: string | null // Express Mode API key
     // Custom headers (e.g., for EdgeOne cookie auth)
     headers?: Record<string, string>
+    // Custom env var names for server models (allows multiple API keys per provider)
+    apiKeyEnv?: string
+    baseUrlEnv?: string
 }
 
 // Providers that can be used with client-provided API keys
@@ -38,6 +44,7 @@ const ALLOWED_CLIENT_PROVIDERS: ProviderName[] = [
     "openai",
     "anthropic",
     "google",
+    "vertexai",
     "azure",
     "bedrock",
     "openrouter",
@@ -88,6 +95,36 @@ export function resolveBaseURL(
     return userBaseUrl || serverBaseUrl || defaultBaseUrl || undefined
 }
 
+/**
+ * Resolve API key from custom env var name or default env var.
+ * Supports multiple API keys per provider via ai-models.json apiKeyEnv config.
+ *
+ * Priority:
+ * 1. User-provided API key (overrides.apiKey)
+ * 2. Custom env var from ai-models.json (overrides.apiKeyEnv)
+ * 3. Default provider env var (defaultEnvVar)
+ */
+function resolveApiKey(
+    overrides: ClientOverrides | undefined,
+    defaultEnvVar: string,
+): string | undefined {
+    if (overrides?.apiKey) return overrides.apiKey
+    if (overrides?.apiKeyEnv) return process.env[overrides.apiKeyEnv]
+    return process.env[defaultEnvVar]
+}
+
+/**
+ * Resolve base URL from custom env var name or default env var.
+ * Supports multiple base URLs per provider via ai-models.json baseUrlEnv config.
+ */
+function resolveBaseUrlEnv(
+    overrides: ClientOverrides | undefined,
+    defaultEnvVar: string,
+): string | undefined {
+    if (overrides?.baseUrlEnv) return process.env[overrides.baseUrlEnv]
+    return process.env[defaultEnvVar]
+}
+
 /**
  * Safely parse integer from environment variable with validation
  */
@@ -122,6 +159,8 @@ function parseIntSafe(
  * - ANTHROPIC_THINKING_TYPE: Anthropic thinking type (enabled)
  * - GOOGLE_THINKING_BUDGET: Google Gemini 2.5 thinking budget in tokens (1024-100000)
  * - GOOGLE_THINKING_LEVEL: Google Gemini 3 thinking level (low/high)
+ * - GOOGLE_VERTEX_THINKING_BUDGET: Vertex AI Gemini 2.5 thinking budget in tokens (1024-100000)
+ * - GOOGLE_VERTEX_THINKING_LEVEL: Vertex AI Gemini 3 thinking level (low/high)
  * - AZURE_REASONING_EFFORT: Azure/OpenAI reasoning effort (low/medium/high)
  * - AZURE_REASONING_SUMMARY: Azure reasoning summary (none/brief/detailed)
  * - BEDROCK_REASONING_BUDGET_TOKENS: Bedrock Claude reasoning budget in tokens (1024-64000)
@@ -286,7 +325,46 @@ function buildProviderOptions(
             }
             break
         }
+        case "vertexai": {
+            const thinkingBudget = parseIntSafe(
+                process.env.GOOGLE_VERTEX_THINKING_BUDGET,
+                "GOOGLE_VERTEX_THINKING_BUDGET",
+                1024,
+                100000,
+            )
+            const thinkingLevel = process.env.GOOGLE_VERTEX_THINKING_LEVEL
 
+            if (
+                modelId &&
+                (modelId.includes("gemini-2") ||
+                    modelId.includes("gemini-3") ||
+                    modelId.includes("gemini2") ||
+                    modelId.includes("gemini3"))
+            ) {
+                const thinkingConfig: Record<string, any> = {
+                    includeThoughts: true,
+                }
+
+                const isGemini3 =
+                    modelId?.includes("gemini-3") ||
+                    modelId?.includes("gemini3")
+                const isGemini25 =
+                    modelId?.includes("2.5") || modelId?.includes("2-5")
+
+                if (isGemini3 && thinkingLevel) {
+                    // Vertex AI provider in AI SDK supports more granular levels (minimal/low/medium/high)
+                    thinkingConfig.thinkingLevel = thinkingLevel as
+                        | "minimal"
+                        | "low"
+                        | "medium"
+                        | "high"
+                } else if (isGemini25 && thinkingBudget) {
+                    thinkingConfig.thinkingBudget = thinkingBudget
+                }
+                options.google = { thinkingConfig }
+            }
+            break
+        }
         case "azure": {
             const reasoningEffort = process.env.AZURE_REASONING_EFFORT
             const reasoningSummary = process.env.AZURE_REASONING_SUMMARY
@@ -388,6 +466,7 @@ const PROVIDER_ENV_VARS: Record<ProviderName, string | null> = {
     openai: "OPENAI_API_KEY",
     anthropic: "ANTHROPIC_API_KEY",
     google: "GOOGLE_GENERATIVE_AI_API_KEY",
+    vertexai: "GOOGLE_VERTEX_API_KEY",
     azure: "AZURE_API_KEY",
     ollama: null, // No credentials needed for local Ollama
     openrouter: "OPENROUTER_API_KEY",
@@ -435,9 +514,15 @@ function detectProvider(): ProviderName | null {
 
 /**
  * Validate that required API keys are present for the selected provider
+ * @param provider - The provider to validate
+ * @param customApiKeyEnv - Optional custom env var name (from ai-models.json apiKeyEnv)
  */
-function validateProviderCredentials(provider: ProviderName): void {
-    const requiredVar = PROVIDER_ENV_VARS[provider]
+function validateProviderCredentials(
+    provider: ProviderName,
+    customApiKeyEnv?: string,
+): void {
+    // Use custom env var name if provided, otherwise use default
+    const requiredVar = customApiKeyEnv || PROVIDER_ENV_VARS[provider]
     if (requiredVar && !process.env[requiredVar]) {
         throw new Error(
             `${requiredVar} environment variable is required for ${provider} provider. ` +
@@ -491,6 +576,7 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
     if (
         overrides?.baseUrl &&
         !overrides?.apiKey &&
+        !(overrides?.provider === "vertexai" && overrides?.vertexApiKey) &&
         overrides?.provider !== "edgeone"
     ) {
         throw new Error(
@@ -500,7 +586,11 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
     }
 
     // Check if client is providing their own provider override
-    const isClientOverride = !!(overrides?.provider && overrides?.apiKey)
+    const isClientOverride = !!(
+        overrides?.provider &&
+        (overrides?.apiKey ||
+            (overrides?.provider === "vertexai" && overrides?.vertexApiKey))
+    )
 
     // Use client override if provided, otherwise fall back to env vars
     const modelId = overrides?.modelId || process.env.AI_MODEL
@@ -570,7 +660,7 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
 
     // Only validate server credentials if client isn't providing their own API key
     if (!isClientOverride) {
-        validateProviderCredentials(provider)
+        validateProviderCredentials(provider, overrides?.apiKeyEnv)
     }
 
     console.log(`[AI Provider] Initializing ${provider} with model: ${modelId}`)
@@ -620,11 +710,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         }
 
         case "openai": {
-            const apiKey = overrides?.apiKey || process.env.OPENAI_API_KEY
+            const apiKey = resolveApiKey(overrides, "OPENAI_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "OPENAI_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.OPENAI_BASE_URL,
+                serverBaseUrl,
             )
             if (baseURL) {
                 // Custom base URL = third-party proxy, use Chat Completions API
@@ -643,11 +737,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         }
 
         case "anthropic": {
-            const apiKey = overrides?.apiKey || process.env.ANTHROPIC_API_KEY
+            const apiKey = resolveApiKey(overrides, "ANTHROPIC_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "ANTHROPIC_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.ANTHROPIC_BASE_URL,
+                serverBaseUrl,
                 "https://api.anthropic.com/v1",
             )
             const customProvider = createAnthropic({
@@ -662,12 +760,18 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         }
 
         case "google": {
-            const apiKey =
-                overrides?.apiKey || process.env.GOOGLE_GENERATIVE_AI_API_KEY
+            const apiKey = resolveApiKey(
+                overrides,
+                "GOOGLE_GENERATIVE_AI_API_KEY",
+            )
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "GOOGLE_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.GOOGLE_BASE_URL,
+                serverBaseUrl,
             )
             if (baseURL || overrides?.apiKey) {
                 const customGoogle = createGoogleGenerativeAI({
@@ -680,13 +784,37 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
             }
             break
         }
+        case "vertexai": {
+            // Express Mode: Use API key for authentication
+            const vertexApiKey =
+                overrides?.vertexApiKey || process.env.GOOGLE_VERTEX_API_KEY
+
+            if (!vertexApiKey) {
+                throw new Error(
+                    "Vertex AI requires an API key for Express Mode. " +
+                        "Get one from Google Cloud Console or set GOOGLE_VERTEX_API_KEY environment variable.",
+                )
+            }
+
+            // Support custom base URL from env or client override
+            const baseURL =
+                overrides?.baseUrl || process.env.GOOGLE_VERTEX_BASE_URL
+
+            const vertexProvider = createVertex({
+                apiKey: vertexApiKey,
+                ...(baseURL && { baseURL }),
+            })
+            model = vertexProvider(modelId)
+            break
+        }
 
         case "azure": {
-            const apiKey = overrides?.apiKey || process.env.AZURE_API_KEY
+            const apiKey = resolveApiKey(overrides, "AZURE_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(overrides, "AZURE_BASE_URL")
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.AZURE_BASE_URL,
+                serverBaseUrl,
             )
             // Only use server's resourceName if user is NOT providing their own API key
             const resourceName = overrides?.apiKey
@@ -720,11 +848,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
             break
 
         case "openrouter": {
-            const apiKey = overrides?.apiKey || process.env.OPENROUTER_API_KEY
+            const apiKey = resolveApiKey(overrides, "OPENROUTER_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "OPENROUTER_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.OPENROUTER_BASE_URL,
+                serverBaseUrl,
             )
             const openrouter = createOpenRouter({
                 apiKey,
@@ -735,11 +867,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         }
 
         case "deepseek": {
-            const apiKey = overrides?.apiKey || process.env.DEEPSEEK_API_KEY
+            const apiKey = resolveApiKey(overrides, "DEEPSEEK_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "DEEPSEEK_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.DEEPSEEK_BASE_URL,
+                serverBaseUrl,
             )
             if (baseURL || overrides?.apiKey) {
                 const customDeepSeek = createDeepSeek({
@@ -754,11 +890,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         }
 
         case "siliconflow": {
-            const apiKey = overrides?.apiKey || process.env.SILICONFLOW_API_KEY
+            const apiKey = resolveApiKey(overrides, "SILICONFLOW_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "SILICONFLOW_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.SILICONFLOW_BASE_URL,
+                serverBaseUrl,
                 "https://api.siliconflow.cn/v1",
             )
             const siliconflowProvider = createOpenAI({
@@ -770,11 +910,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         }
 
         case "sglang": {
-            const apiKey = overrides?.apiKey || process.env.SGLANG_API_KEY
+            const apiKey = resolveApiKey(overrides, "SGLANG_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "SGLANG_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.SGLANG_BASE_URL,
+                serverBaseUrl,
             )
 
             const sglangProvider = createOpenAI({
@@ -883,11 +1027,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
             // Vercel AI Gateway - unified access to multiple AI providers
             // Model format: "provider/model" e.g., "openai/gpt-4o", "anthropic/claude-sonnet-4-5"
             // See: https://vercel.com/ai-gateway
-            const apiKey = overrides?.apiKey || process.env.AI_GATEWAY_API_KEY
+            const apiKey = resolveApiKey(overrides, "AI_GATEWAY_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "AI_GATEWAY_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.AI_GATEWAY_BASE_URL,
+                serverBaseUrl,
             )
             // Only use custom configuration if explicitly set (local dev or custom Gateway)
             // Otherwise undefined → AI SDK uses Vercel default (https://ai-gateway.vercel.sh/v1/ai) + OIDC
@@ -919,11 +1067,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         }
 
         case "doubao": {
-            const apiKey = overrides?.apiKey || process.env.DOUBAO_API_KEY
+            const apiKey = resolveApiKey(overrides, "DOUBAO_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "DOUBAO_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.DOUBAO_BASE_URL,
+                serverBaseUrl,
                 "https://ark.cn-beijing.volces.com/api/v3",
             )
             const lowerModelId = modelId.toLowerCase()
@@ -948,11 +1100,15 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
         }
 
         case "modelscope": {
-            const apiKey = overrides?.apiKey || process.env.MODELSCOPE_API_KEY
+            const apiKey = resolveApiKey(overrides, "MODELSCOPE_API_KEY")
+            const serverBaseUrl = resolveBaseUrlEnv(
+                overrides,
+                "MODELSCOPE_BASE_URL",
+            )
             const baseURL = resolveBaseURL(
                 overrides?.apiKey,
                 overrides?.baseUrl,
-                process.env.MODELSCOPE_BASE_URL,
+                serverBaseUrl,
                 "https://api-inference.modelscope.cn/v1",
             )
             const modelscopeProvider = createOpenAI({

lib/i18n/dictionaries/en.json

@@ -305,10 +305,13 @@
         "noModelsFound": "No models found.",
         "default": "Default",
         "serverDefault": "Server Default",
+        "serverModels": "Server Models",
+        "userModels": "User Models",
         "configureModels": "Configure Models...",
         "onlyVerifiedShown": "Only verified models are shown",
         "showUnvalidatedModels": "Show unvalidated models",
         "allModelsShown": "All models are shown (including unvalidated)",
-        "unvalidatedModelWarning": "This model has not been validated"
+        "unvalidatedModelWarning": "This model has not been validated",
+        "serverDefaultModel": "Server default model"
     }
 }

lib/i18n/dictionaries/ja.json

@@ -305,10 +305,13 @@
         "noModelsFound": "モデルが見つかりません。",
         "default": "デフォルト",
         "serverDefault": "サーバーデフォルト",
+        "serverModels": "サーバーモデル",
+        "userModels": "ユーザーモデル",
         "configureModels": "モデルを設定...",
         "onlyVerifiedShown": "検証済みのモデルのみ表示",
         "showUnvalidatedModels": "未検証のモデルを表示",
         "allModelsShown": "すべてのモデルを表示（未検証を含む）",
-        "unvalidatedModelWarning": "このモデルは検証されていません"
+        "unvalidatedModelWarning": "このモデルは検証されていません",
+        "serverDefaultModel": "サーバーデフォルトモデル"
     }
 }

lib/i18n/dictionaries/zh.json

@@ -305,10 +305,13 @@
         "noModelsFound": "未找到模型。",
         "default": "默认",
         "serverDefault": "服务器默认",
+        "serverModels": "服务器模型",
+        "userModels": "用户模型",
         "configureModels": "配置模型...",
         "onlyVerifiedShown": "仅显示已验证的模型",
         "showUnvalidatedModels": "显示未验证的模型",
         "allModelsShown": "显示所有模型（包括未验证的）",
-        "unvalidatedModelWarning": "此模型尚未验证"
+        "unvalidatedModelWarning": "此模型尚未验证",
+        "serverDefaultModel": "服务器默认模型"
     }
 }

lib/server-model-config.ts

@@ -0,0 +1,151 @@
+import fs from "fs/promises"
+import path from "path"
+import { z } from "zod"
+import type { ProviderName } from "@/lib/types/model-config"
+import { PROVIDER_INFO } from "@/lib/types/model-config"
+
+export const ProviderNameSchema: z.ZodType<ProviderName> = z
+    .string()
+    .refine((val): val is ProviderName => val in PROVIDER_INFO, {
+        message: "Invalid provider name",
+    })
+
+export const ServerProviderSchema = z.object({
+    name: z.string().min(1),
+    provider: ProviderNameSchema,
+    models: z.array(z.string().min(1)),
+    // Optional: custom environment variable name for API key
+    // e.g., "OPENAI_API_KEY_TEAM_A" instead of default "OPENAI_API_KEY"
+    apiKeyEnv: z.string().min(1).optional(),
+    // Optional: custom environment variable name for base URL
+    baseUrlEnv: z.string().min(1).optional(),
+    // Optional: mark the first model in this provider as the default
+    default: z.boolean().optional(),
+})
+
+export const ServerModelsConfigSchema = z.object({
+    providers: z.array(ServerProviderSchema),
+})
+
+export type ServerProviderConfig = z.infer<typeof ServerProviderSchema>
+export type ServerModelsConfig = z.infer<typeof ServerModelsConfigSchema>
+
+export interface FlattenedServerModel {
+    id: string // "server:<slugified-name>:<modelId>" - name ensures uniqueness for multiple API keys per provider
+    modelId: string
+    provider: ProviderName
+    providerLabel: string
+    isDefault: boolean
+    // Custom env var names for credentials (optional)
+    apiKeyEnv?: string
+    baseUrlEnv?: string
+}
+
+/**
+ * Convert provider name to URL-safe slug for use in model ID
+ * e.g., "OpenAI Production" → "openai-production"
+ */
+function slugify(name: string): string {
+    return name
+        .toLowerCase()
+        .replace(/[^a-z0-9]+/g, "-")
+        .replace(/^-|-$/g, "")
+}
+
+function getConfigPath(): string {
+    const custom = process.env.AI_MODELS_CONFIG_PATH
+    if (custom && custom.trim().length > 0) return custom
+    return path.join(process.cwd(), "ai-models.json")
+}
+
+export async function loadRawServerModelsConfig(): Promise<ServerModelsConfig | null> {
+    // Priority 1: AI_MODELS_CONFIG env var (JSON string) - for cloud deployments
+    const envConfig = process.env.AI_MODELS_CONFIG
+    if (envConfig && envConfig.trim().length > 0) {
+        try {
+            const json = JSON.parse(envConfig)
+            return ServerModelsConfigSchema.parse(json)
+        } catch (err) {
+            console.error(
+                "[server-model-config] Failed to parse AI_MODELS_CONFIG:",
+                err,
+            )
+            return null
+        }
+    }
+
+    // Priority 2: ai-models.json file
+    const configPath = getConfigPath()
+    try {
+        const jsonStr = await fs.readFile(configPath, "utf8")
+        const json = JSON.parse(jsonStr)
+        return ServerModelsConfigSchema.parse(json)
+    } catch (err: any) {
+        if (err?.code === "ENOENT") {
+            return null
+        }
+        console.error(
+            "[server-model-config] Failed to load ai-models.json:",
+            err,
+        )
+        return null
+    }
+}
+
+export async function loadFlattenedServerModels(): Promise<
+    FlattenedServerModel[]
+> {
+    const cfg = await loadRawServerModelsConfig()
+    if (!cfg) return []
+
+    const defaultProvider = process.env.AI_PROVIDER as ProviderName | undefined
+    const defaultModelId = process.env.AI_MODEL
+
+    const flattened: FlattenedServerModel[] = []
+
+    for (const p of cfg.providers) {
+        const providerLabel =
+            p.name || PROVIDER_INFO[p.provider]?.label || p.provider
+
+        // Use slugified name for unique ID (supports multiple API keys per provider)
+        const nameSlug = slugify(p.name)
+
+        for (const modelId of p.models) {
+            const id = `server:${nameSlug}:${modelId}`
+
+            // Default model priority:
+            // 1. From ai-models.json: first model of provider with default: true
+            // 2. From env vars: AI_MODEL matches (legacy behavior)
+            const isDefault =
+                (p.default === true && modelId === p.models[0]) ||
+                (!!defaultModelId &&
+                    modelId === defaultModelId &&
+                    (!defaultProvider || defaultProvider === p.provider))
+
+            flattened.push({
+                id,
+                modelId,
+                provider: p.provider,
+                providerLabel,
+                isDefault,
+                apiKeyEnv: p.apiKeyEnv,
+                baseUrlEnv: p.baseUrlEnv,
+            })
+        }
+    }
+
+    return flattened
+}
+
+/**
+ * Find a server model by its ID (format: "server:<slugified-name>:<modelId>")
+ * Returns the model config including apiKeyEnv/baseUrlEnv if configured
+ */
+export async function findServerModelById(
+    modelId: string,
+): Promise<FlattenedServerModel | null> {
+    if (!modelId.startsWith("server:")) return null
+
+    const models = await loadFlattenedServerModels()
+    return models.find((m) => m.id === modelId) || null
+}

lib/types/model-config.ts

@@ -4,6 +4,7 @@ export type ProviderName =
     | "openai"
     | "anthropic"
     | "google"
+    | "vertexai"
     | "azure"
     | "bedrock"
     | "ollama"
@@ -36,6 +37,9 @@ export interface ProviderConfig {
     awsSecretAccessKey?: string
     awsRegion?: string
     awsSessionToken?: string // Optional, for temporary credentials
+    // Vertex AI specific fields
+    vertexApiKey?: string // Express Mode API key
+
     models: ModelConfig[]
     validated?: boolean // Has API key been validated
 }
@@ -50,7 +54,7 @@ export interface MultiModelConfig {
 
 // Flattened model for dropdown display
 export interface FlattenedModel {
-    id: string // Model config UUID
+    id: string // Model config UUID or synthetic server ID (e.g., "server:provider:modelId")
     modelId: string // Actual model ID
     provider: ProviderName
     providerLabel: string // Provider display name
@@ -61,7 +65,17 @@ export interface FlattenedModel {
     awsSecretAccessKey?: string
     awsRegion?: string
     awsSessionToken?: string
+    // Vertex AI specific fields
+    vertexApiKey?: string // Express Mode API key
+
     validated?: boolean // Has this model been validated
+    // Source of this model config: user-defined (client) or server-defined
+    source?: "user" | "server"
+    // Whether this model is the server default (matches AI_MODEL env var)
+    isDefault?: boolean
+    // Custom env var names for server models (allows multiple API keys per provider)
+    apiKeyEnv?: string
+    baseUrlEnv?: string
 }
 
 // Provider metadata
@@ -75,6 +89,7 @@ export const PROVIDER_INFO: Record<
         defaultBaseUrl: "https://api.anthropic.com/v1",
     },
     google: { label: "Google" },
+    vertexai: { label: "Google Vertex AI" },
     azure: { label: "Azure OpenAI" },
     bedrock: { label: "Amazon Bedrock" },
     ollama: {
@@ -157,6 +172,17 @@ export const SUGGESTED_MODELS: Partial<Record<ProviderName, string[]>> = {
         // Legacy
         "gemini-pro",
     ],
+    vertexai: [
+        // Gemini 2.5 series
+        "gemini-2.5-pro",
+        "gemini-2.5-flash",
+        // Gemini 2.0 series
+        "gemini-2.0-flash",
+        "gemini-2.0-flash-exp",
+        // Gemini 1.5 series
+        "gemini-1.5-pro",
+        "gemini-1.5-flash",
+    ],
     azure: ["gpt-4o", "gpt-4o-mini", "gpt-4-turbo", "gpt-4", "gpt-35-turbo"],
     bedrock: [
         // Anthropic Claude
@@ -288,7 +314,7 @@ export function createModelConfig(modelId: string): ModelConfig {
     }
 }
 
-// Get all models as flattened list for dropdown
+// Get all models as flattened list for dropdown (user-defined only)
 export function flattenModels(config: MultiModelConfig): FlattenedModel[] {
     const models: FlattenedModel[] = []
 
@@ -310,7 +336,12 @@ export function flattenModels(config: MultiModelConfig): FlattenedModel[] {
                 awsSecretAccessKey: provider.awsSecretAccessKey,
                 awsRegion: provider.awsRegion,
                 awsSessionToken: provider.awsSessionToken,
+                // Vertex AI fields
+                vertexApiKey: provider.vertexApiKey,
+
                 validated: model.validated,
+                source: "user",
+                isDefault: false,
             })
         }
     }

package-lock.json

@@ -15,6 +15,7 @@
                 "@ai-sdk/deepseek": "^2.0.0",
                 "@ai-sdk/gateway": "^3.0.0",
                 "@ai-sdk/google": "^3.0.0",
+                "@ai-sdk/google-vertex": "^4.0.16",
                 "@ai-sdk/openai": "^3.0.0",
                 "@ai-sdk/react": "^3.0.1",
                 "@aws-sdk/client-dynamodb": "^3.957.0",
@@ -206,13 +207,106 @@
             }
         },
         "node_modules/@ai-sdk/google": {
-            "version": "3.0.6",
-            "resolved": "https://registry.npmjs.org/@ai-sdk/google/-/google-3.0.6.tgz",
-            "integrity": "sha512-Nr7E+ouWd/bKO9SFlgLnJJ1+fiGHC07KAeFr08faT+lvkECWlxVox3aL0dec8uCgBDUghYbq7f4S5teUrCc+QQ==",
+            "version": "3.0.9",
+            "resolved": "https://registry.npmjs.org/@ai-sdk/google/-/google-3.0.9.tgz",
+            "integrity": "sha512-whRdK0gCZL92UbEHdmQ6edm3cp5ZZSqwE79AIvTEaJR+BeaMUJchTxz/I5w9l3EHGOiDxrKYssklqO3z45KGXg==",
             "license": "Apache-2.0",
             "dependencies": {
-                "@ai-sdk/provider": "3.0.2",
-                "@ai-sdk/provider-utils": "4.0.4"
+                "@ai-sdk/provider": "3.0.3",
+                "@ai-sdk/provider-utils": "4.0.7"
+            },
+            "engines": {
+                "node": ">=18"
+            },
+            "peerDependencies": {
+                "zod": "^3.25.76 || ^4.1.8"
+            }
+        },
+        "node_modules/@ai-sdk/google-vertex": {
+            "version": "4.0.16",
+            "resolved": "https://registry.npmjs.org/@ai-sdk/google-vertex/-/google-vertex-4.0.16.tgz",
+            "integrity": "sha512-paY4NGCaqMbd0kssK27ssTMggDybNPIFudJ99QsdzRDJ/Kgo0EjeXZn7G/CoBY/Snmf0fA0PEZW34yV9kloFCQ==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "@ai-sdk/anthropic": "3.0.14",
+                "@ai-sdk/google": "3.0.9",
+                "@ai-sdk/provider": "3.0.3",
+                "@ai-sdk/provider-utils": "4.0.7",
+                "google-auth-library": "^10.5.0"
+            },
+            "engines": {
+                "node": ">=18"
+            },
+            "peerDependencies": {
+                "zod": "^3.25.76 || ^4.1.8"
+            }
+        },
+        "node_modules/@ai-sdk/google-vertex/node_modules/@ai-sdk/anthropic": {
+            "version": "3.0.14",
+            "resolved": "https://registry.npmjs.org/@ai-sdk/anthropic/-/anthropic-3.0.14.tgz",
+            "integrity": "sha512-71BaVg60FM6tN0JaRY7kRb/6qZtHi5R9PFF3NES+kqonY3nVD4PCPP8DoMb/tqxC7f/XezlCqsHrveT2mGfi3A==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "@ai-sdk/provider": "3.0.3",
+                "@ai-sdk/provider-utils": "4.0.7"
+            },
+            "engines": {
+                "node": ">=18"
+            },
+            "peerDependencies": {
+                "zod": "^3.25.76 || ^4.1.8"
+            }
+        },
+        "node_modules/@ai-sdk/google-vertex/node_modules/@ai-sdk/provider": {
+            "version": "3.0.3",
+            "resolved": "https://registry.npmjs.org/@ai-sdk/provider/-/provider-3.0.3.tgz",
+            "integrity": "sha512-qGPYdoAuECaUXPrrz0BPX1SacZQuJ6zky0aakxpW89QW1hrY0eF4gcFm/3L9Pk8C5Fwe+RvBf2z7ZjDhaPjnlg==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "json-schema": "^0.4.0"
+            },
+            "engines": {
+                "node": ">=18"
+            }
+        },
+        "node_modules/@ai-sdk/google-vertex/node_modules/@ai-sdk/provider-utils": {
+            "version": "4.0.7",
+            "resolved": "https://registry.npmjs.org/@ai-sdk/provider-utils/-/provider-utils-4.0.7.tgz",
+            "integrity": "sha512-ItzTdBxRLieGz1GHPwl9X3+HKfwTfFd9MdIa91aXRnOjUVRw68ENjAGKm3FcXGsBLkXDLaFWgjbTVdXe2igs2w==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "@ai-sdk/provider": "3.0.3",
+                "@standard-schema/spec": "^1.1.0",
+                "eventsource-parser": "^3.0.6"
+            },
+            "engines": {
+                "node": ">=18"
+            },
+            "peerDependencies": {
+                "zod": "^3.25.76 || ^4.1.8"
+            }
+        },
+        "node_modules/@ai-sdk/google/node_modules/@ai-sdk/provider": {
+            "version": "3.0.3",
+            "resolved": "https://registry.npmjs.org/@ai-sdk/provider/-/provider-3.0.3.tgz",
+            "integrity": "sha512-qGPYdoAuECaUXPrrz0BPX1SacZQuJ6zky0aakxpW89QW1hrY0eF4gcFm/3L9Pk8C5Fwe+RvBf2z7ZjDhaPjnlg==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "json-schema": "^0.4.0"
+            },
+            "engines": {
+                "node": ">=18"
+            }
+        },
+        "node_modules/@ai-sdk/google/node_modules/@ai-sdk/provider-utils": {
+            "version": "4.0.7",
+            "resolved": "https://registry.npmjs.org/@ai-sdk/provider-utils/-/provider-utils-4.0.7.tgz",
+            "integrity": "sha512-ItzTdBxRLieGz1GHPwl9X3+HKfwTfFd9MdIa91aXRnOjUVRw68ENjAGKm3FcXGsBLkXDLaFWgjbTVdXe2igs2w==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "@ai-sdk/provider": "3.0.3",
+                "@standard-schema/spec": "^1.1.0",
+                "eventsource-parser": "^3.0.6"
             },
             "engines": {
                 "node": ">=18"
@@ -9768,7 +9862,6 @@
             "version": "0.11.0",
             "resolved": "https://registry.npmjs.org/@pkgjs/parseargs/-/parseargs-0.11.0.tgz",
             "integrity": "sha512-+1VkjdD0QBLPodGrJUeqarH8VAIvQODIbwh9XpP5Syisf7YoQgsJKPNFoqqLQlu+VQ/tVSshMR6loPMn8U+dPg==",
-            "dev": true,
             "license": "MIT",
             "optional": true,
             "engines": {
@@ -13467,7 +13560,6 @@
             "version": "7.1.3",
             "resolved": "https://registry.npmjs.org/agent-base/-/agent-base-7.1.3.tgz",
             "integrity": "sha512-jRR5wdylq8CkOe6hei19GGZnxM6rBGwFl3Bg0YItGDimvjGtAvdZk4Pu6Cl4u4Igsws4a1fd1Vq3ezrhn4KmFw==",
-            "dev": true,
             "license": "MIT",
             "engines": {
                 "node": ">= 14"
@@ -14091,7 +14183,6 @@
             "version": "1.5.1",
             "resolved": "https://registry.npmjs.org/base64-js/-/base64-js-1.5.1.tgz",
             "integrity": "sha512-AKpaYlHn8t4SVbOHCy+b5+KKgvR4vrsD8vbvrbiQJps7fKDTkjkDry6ji0rUJjC0kzbNePLwzxq8iypo41qeWA==",
-            "dev": true,
             "funding": [
                 {
                     "type": "github",
@@ -14127,6 +14218,15 @@
                 "require-from-string": "^2.0.2"
             }
         },
+        "node_modules/bignumber.js": {
+            "version": "9.3.1",
+            "resolved": "https://registry.npmjs.org/bignumber.js/-/bignumber.js-9.3.1.tgz",
+            "integrity": "sha512-Ko0uX15oIUS7wJ3Rb30Fs6SkVbLmPBAKdlm7q9+ak9bbIeFf0MwuBsQV6z7+X768/cHsfg+WlysDWJcmthjsjQ==",
+            "license": "MIT",
+            "engines": {
+                "node": "*"
+            }
+        },
         "node_modules/bl": {
             "version": "4.1.0",
             "resolved": "https://registry.npmjs.org/bl/-/bl-4.1.0.tgz",
@@ -14300,6 +14400,12 @@
                 "node": "*"
             }
         },
+        "node_modules/buffer-equal-constant-time": {
+            "version": "1.0.1",
+            "resolved": "https://registry.npmjs.org/buffer-equal-constant-time/-/buffer-equal-constant-time-1.0.1.tgz",
+            "integrity": "sha512-zRpUiDwd/xk6ADqPMATG8vc9VPrkck7T07OIx0gnjmJAnHnTVXNQG3vfvWNuiZIkwu9KrKdA1iJKfsfTVxE6NA==",
+            "license": "BSD-3-Clause"
+        },
         "node_modules/buffer-from": {
             "version": "1.1.2",
             "resolved": "https://registry.npmmirror.com/buffer-from/-/buffer-from-1.1.2.tgz",
@@ -15293,6 +15399,15 @@
             "dev": true,
             "license": "BSD-2-Clause"
         },
+        "node_modules/data-uri-to-buffer": {
+            "version": "4.0.1",
+            "resolved": "https://registry.npmjs.org/data-uri-to-buffer/-/data-uri-to-buffer-4.0.1.tgz",
+            "integrity": "sha512-0R9ikRb668HB7QDxT1vkpuUBtqc53YyAwMwGeUFKRojY/NWKvdZ+9UYtRfGmhqNbRkTSVpMbmyhXipFFv2cb/A==",
+            "license": "MIT",
+            "engines": {
+                "node": ">= 12"
+            }
+        },
         "node_modules/data-urls": {
             "version": "6.0.0",
             "resolved": "https://registry.npmjs.org/data-urls/-/data-urls-6.0.0.tgz",
@@ -15800,6 +15915,15 @@
             "integrity": "sha512-I88TYZWc9XiYHRQ4/3c5rjjfgkjhLyW2luGIheGERbNQ6OY7yTybanSpDXZa8y7VUP9YmDcYa+eyq4ca7iLqWA==",
             "license": "MIT"
         },
+        "node_modules/ecdsa-sig-formatter": {
+            "version": "1.0.11",
+            "resolved": "https://registry.npmjs.org/ecdsa-sig-formatter/-/ecdsa-sig-formatter-1.0.11.tgz",
+            "integrity": "sha512-nagl3RYrbNv6kQkeJIpt6NJZy8twLB/2vtz6yN9Z4vRKHN4/QZJIEbqohALSgwKdnksuY3k5Addp5lg8sVoVcQ==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "safe-buffer": "^5.0.1"
+            }
+        },
         "node_modules/eciesjs": {
             "version": "0.4.16",
             "resolved": "https://registry.npmjs.org/eciesjs/-/eciesjs-0.4.16.tgz",
@@ -17282,6 +17406,38 @@
                 "pend": "~1.2.0"
             }
         },
+        "node_modules/fetch-blob": {
+            "version": "3.2.0",
+            "resolved": "https://registry.npmjs.org/fetch-blob/-/fetch-blob-3.2.0.tgz",
+            "integrity": "sha512-7yAQpD2UMJzLi1Dqv7qFYnPbaPx7ZfFK6PiIxQ4PfkGPyNyl2Ugx+a/umUonmKqjhM4DnfbMvdX6otXq83soQQ==",
+            "funding": [
+                {
+                    "type": "github",
+                    "url": "https://github.com/sponsors/jimmywarting"
+                },
+                {
+                    "type": "paypal",
+                    "url": "https://paypal.me/jimmywarting"
+                }
+            ],
+            "license": "MIT",
+            "dependencies": {
+                "node-domexception": "^1.0.0",
+                "web-streams-polyfill": "^3.0.3"
+            },
+            "engines": {
+                "node": "^12.20 || >= 14.13"
+            }
+        },
+        "node_modules/fetch-blob/node_modules/web-streams-polyfill": {
+            "version": "3.3.3",
+            "resolved": "https://registry.npmjs.org/web-streams-polyfill/-/web-streams-polyfill-3.3.3.tgz",
+            "integrity": "sha512-d2JWLCivmZYTSIoge9MsgFCZrt571BikcWGYkjC1khllbTeDlGqZ2D8vD8E/lJa8WGWbb7Plm8/XJYV7IJHZZw==",
+            "license": "MIT",
+            "engines": {
+                "node": ">= 8"
+            }
+        },
         "node_modules/file-entry-cache": {
             "version": "8.0.0",
             "resolved": "https://registry.npmjs.org/file-entry-cache/-/file-entry-cache-8.0.0.tgz",
@@ -17488,6 +17644,18 @@
                 "node": ">= 12.20"
             }
         },
+        "node_modules/formdata-polyfill": {
+            "version": "4.0.10",
+            "resolved": "https://registry.npmjs.org/formdata-polyfill/-/formdata-polyfill-4.0.10.tgz",
+            "integrity": "sha512-buewHzMvYL29jdeQTVILecSaZKnt/RJWjoZCF5OW60Z67/GmSLBkOFM7qh1PI3zFNtJbaZL5eQu1vLfazOwj4g==",
+            "license": "MIT",
+            "dependencies": {
+                "fetch-blob": "^3.1.2"
+            },
+            "engines": {
+                "node": ">=12.20.0"
+            }
+        },
         "node_modules/forwarded": {
             "version": "0.2.0",
             "resolved": "https://registry.npmjs.org/forwarded/-/forwarded-0.2.0.tgz",
@@ -17621,6 +17789,112 @@
                 "url": "https://github.com/sponsors/ljharb"
             }
         },
+        "node_modules/gaxios": {
+            "version": "7.1.3",
+            "resolved": "https://registry.npmjs.org/gaxios/-/gaxios-7.1.3.tgz",
+            "integrity": "sha512-YGGyuEdVIjqxkxVH1pUTMY/XtmmsApXrCVv5EU25iX6inEPbV+VakJfLealkBtJN69AQmh1eGOdCl9Sm1UP6XQ==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "extend": "^3.0.2",
+                "https-proxy-agent": "^7.0.1",
+                "node-fetch": "^3.3.2",
+                "rimraf": "^5.0.1"
+            },
+            "engines": {
+                "node": ">=18"
+            }
+        },
+        "node_modules/gaxios/node_modules/brace-expansion": {
+            "version": "2.0.2",
+            "resolved": "https://registry.npmjs.org/brace-expansion/-/brace-expansion-2.0.2.tgz",
+            "integrity": "sha512-Jt0vHyM+jmUBqojB7E1NIYadt0vI0Qxjxd2TErW94wDz+E2LAm5vKMXXwg6ZZBTHPuUlDgQHKXvjGBdfcF1ZDQ==",
+            "license": "MIT",
+            "dependencies": {
+                "balanced-match": "^1.0.0"
+            }
+        },
+        "node_modules/gaxios/node_modules/glob": {
+            "version": "10.5.0",
+            "resolved": "https://registry.npmjs.org/glob/-/glob-10.5.0.tgz",
+            "integrity": "sha512-DfXN8DfhJ7NH3Oe7cFmu3NCu1wKbkReJ8TorzSAFbSKrlNaQSKfIzqYqVY8zlbs2NLBbWpRiU52GX2PbaBVNkg==",
+            "license": "ISC",
+            "dependencies": {
+                "foreground-child": "^3.1.0",
+                "jackspeak": "^3.1.2",
+                "minimatch": "^9.0.4",
+                "minipass": "^7.1.2",
+                "package-json-from-dist": "^1.0.0",
+                "path-scurry": "^1.11.1"
+            },
+            "bin": {
+                "glob": "dist/esm/bin.mjs"
+            },
+            "funding": {
+                "url": "https://github.com/sponsors/isaacs"
+            }
+        },
+        "node_modules/gaxios/node_modules/minimatch": {
+            "version": "9.0.5",
+            "resolved": "https://registry.npmjs.org/minimatch/-/minimatch-9.0.5.tgz",
+            "integrity": "sha512-G6T0ZX48xgozx7587koeX9Ys2NYy6Gmv//P89sEte9V9whIapMNF4idKxnW2QtCcLiTWlb/wfCabAtAFWhhBow==",
+            "license": "ISC",
+            "dependencies": {
+                "brace-expansion": "^2.0.1"
+            },
+            "engines": {
+                "node": ">=16 || 14 >=14.17"
+            },
+            "funding": {
+                "url": "https://github.com/sponsors/isaacs"
+            }
+        },
+        "node_modules/gaxios/node_modules/node-fetch": {
+            "version": "3.3.2",
+            "resolved": "https://registry.npmjs.org/node-fetch/-/node-fetch-3.3.2.tgz",
+            "integrity": "sha512-dRB78srN/l6gqWulah9SrxeYnxeddIG30+GOqK/9OlLVyLg3HPnr6SqOWTWOXKRwC2eGYCkZ59NNuSgvSrpgOA==",
+            "license": "MIT",
+            "dependencies": {
+                "data-uri-to-buffer": "^4.0.0",
+                "fetch-blob": "^3.1.4",
+                "formdata-polyfill": "^4.0.10"
+            },
+            "engines": {
+                "node": "^12.20.0 || ^14.13.1 || >=16.0.0"
+            },
+            "funding": {
+                "type": "opencollective",
+                "url": "https://opencollective.com/node-fetch"
+            }
+        },
+        "node_modules/gaxios/node_modules/rimraf": {
+            "version": "5.0.10",
+            "resolved": "https://registry.npmjs.org/rimraf/-/rimraf-5.0.10.tgz",
+            "integrity": "sha512-l0OE8wL34P4nJH/H2ffoaniAokM2qSmrtXHmlpvYr5AVVX8msAyW0l8NVJFDxlSK4u3Uh/f41cQheDVdnYijwQ==",
+            "license": "ISC",
+            "dependencies": {
+                "glob": "^10.3.7"
+            },
+            "bin": {
+                "rimraf": "dist/esm/bin.mjs"
+            },
+            "funding": {
+                "url": "https://github.com/sponsors/isaacs"
+            }
+        },
+        "node_modules/gcp-metadata": {
+            "version": "8.1.2",
+            "resolved": "https://registry.npmjs.org/gcp-metadata/-/gcp-metadata-8.1.2.tgz",
+            "integrity": "sha512-zV/5HKTfCeKWnxG0Dmrw51hEWFGfcF2xiXqcA3+J90WDuP0SvoiSO5ORvcBsifmx/FoIjgQN3oNOGaQ5PhLFkg==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "gaxios": "^7.0.0",
+                "google-logging-utils": "^1.0.0",
+                "json-bigint": "^1.0.0"
+            },
+            "engines": {
+                "node": ">=18"
+            }
+        },
         "node_modules/generator-function": {
             "version": "2.0.1",
             "resolved": "https://registry.npmjs.org/generator-function/-/generator-function-2.0.1.tgz",
@@ -17852,6 +18126,33 @@
             "dev": true,
             "license": "MIT"
         },
+        "node_modules/google-auth-library": {
+            "version": "10.5.0",
+            "resolved": "https://registry.npmjs.org/google-auth-library/-/google-auth-library-10.5.0.tgz",
+            "integrity": "sha512-7ABviyMOlX5hIVD60YOfHw4/CxOfBhyduaYB+wbFWCWoni4N7SLcV46hrVRktuBbZjFC9ONyqamZITN7q3n32w==",
+            "license": "Apache-2.0",
+            "dependencies": {
+                "base64-js": "^1.3.0",
+                "ecdsa-sig-formatter": "^1.0.11",
+                "gaxios": "^7.0.0",
+                "gcp-metadata": "^8.0.0",
+                "google-logging-utils": "^1.0.0",
+                "gtoken": "^8.0.0",
+                "jws": "^4.0.0"
+            },
+            "engines": {
+                "node": ">=18"
+            }
+        },
+        "node_modules/google-logging-utils": {
+            "version": "1.1.3",
+            "resolved": "https://registry.npmjs.org/google-logging-utils/-/google-logging-utils-1.1.3.tgz",
+            "integrity": "sha512-eAmLkjDjAFCVXg7A1unxHsLf961m6y17QFqXqAXGj/gVkKFrEICfStRfwUlGNfeCEjNRa32JEWOUTlYXPyyKvA==",
+            "license": "Apache-2.0",
+            "engines": {
+                "node": ">=14"
+            }
+        },
         "node_modules/gopd": {
             "version": "1.2.0",
             "resolved": "https://registry.npmjs.org/gopd/-/gopd-1.2.0.tgz",
@@ -17904,6 +18205,19 @@
             "dev": true,
             "license": "MIT"
         },
+        "node_modules/gtoken": {
+            "version": "8.0.0",
+            "resolved": "https://registry.npmjs.org/gtoken/-/gtoken-8.0.0.tgz",
+            "integrity": "sha512-+CqsMbHPiSTdtSO14O51eMNlrp9N79gmeqmXeouJOhfucAedHw9noVe/n5uJk3tbKE6a+6ZCQg3RPhVhHByAIw==",
+            "license": "MIT",
+            "dependencies": {
+                "gaxios": "^7.0.0",
+                "jws": "^4.0.0"
+            },
+            "engines": {
+                "node": ">=18"
+            }
+        },
         "node_modules/gzip-size": {
             "version": "6.0.0",
             "resolved": "https://registry.npmjs.org/gzip-size/-/gzip-size-6.0.0.tgz",
@@ -18208,7 +18522,6 @@
             "version": "7.0.6",
             "resolved": "https://registry.npmjs.org/https-proxy-agent/-/https-proxy-agent-7.0.6.tgz",
             "integrity": "sha512-vK9P5/iUfdl95AI+JVyUuIcVtd4ofvtrOr3HNtM2yxC9bnMbEdp3x01OhQNnjb8IJYi38VlTE3mBXwcfvywuSw==",
-            "dev": true,
             "license": "MIT",
             "dependencies": {
                 "agent-base": "^7.1.2",
@@ -19145,7 +19458,6 @@
             "version": "3.4.3",
             "resolved": "https://registry.npmjs.org/jackspeak/-/jackspeak-3.4.3.tgz",
             "integrity": "sha512-OGlZQpz2yfahA/Rd1Y8Cd9SIEsqvXkLVoSw/cgwhnhFMDbsQFeZYoJJ7bIZBS9BcamUW96asq/npPWugM+RQBw==",
-            "dev": true,
             "license": "BlueOak-1.0.0",
             "dependencies": {
                 "@isaacs/cliui": "^8.0.2"
@@ -19278,6 +19590,15 @@
                 "node": ">=6"
             }
         },
+        "node_modules/json-bigint": {
+            "version": "1.0.0",
+            "resolved": "https://registry.npmjs.org/json-bigint/-/json-bigint-1.0.0.tgz",
+            "integrity": "sha512-SiPv/8VpZuWbvLSMtTDU8hEfrZWg/mH/nV/b4o0CYbSxu1UIQPLdwKOCIyLQX+VIPO5vrLX3i8qtqFyhdPSUSQ==",
+            "license": "MIT",
+            "dependencies": {
+                "bignumber.js": "^9.0.0"
+            }
+        },
         "node_modules/json-buffer": {
             "version": "3.0.1",
             "resolved": "https://registry.npmjs.org/json-buffer/-/json-buffer-3.0.1.tgz",
@@ -19361,6 +19682,27 @@
                 "node": ">=4.0"
             }
         },
+        "node_modules/jwa": {
+            "version": "2.0.1",
+            "resolved": "https://registry.npmjs.org/jwa/-/jwa-2.0.1.tgz",
+            "integrity": "sha512-hRF04fqJIP8Abbkq5NKGN0Bbr3JxlQ+qhZufXVr0DvujKy93ZCbXZMHDL4EOtodSbCWxOqR8MS1tXA5hwqCXDg==",
+            "license": "MIT",
+            "dependencies": {
+                "buffer-equal-constant-time": "^1.0.1",
+                "ecdsa-sig-formatter": "1.0.11",
+                "safe-buffer": "^5.0.1"
+            }
+        },
+        "node_modules/jws": {
+            "version": "4.0.1",
+            "resolved": "https://registry.npmjs.org/jws/-/jws-4.0.1.tgz",
+            "integrity": "sha512-EKI/M/yqPncGUUh44xz0PxSidXFr/+r0pA70+gIYhjv+et7yxM+s29Y+VGDkovRofQem0fs7Uvf4+YmAdyRduA==",
+            "license": "MIT",
+            "dependencies": {
+                "jwa": "^2.0.1",
+                "safe-buffer": "^5.0.1"
+            }
+        },
         "node_modules/keyv": {
             "version": "4.5.4",
             "resolved": "https://registry.npmjs.org/keyv/-/keyv-4.5.4.tgz",
@@ -23742,7 +24084,6 @@
             "version": "5.2.1",
             "resolved": "https://registry.npmjs.org/safe-buffer/-/safe-buffer-5.2.1.tgz",
             "integrity": "sha512-rp3So07KcdmmKbGvgaNxQSJr7bGVSVk5S9Eq1F+ppbRo70+YeaDxkw5Dd8NPN+GD6bjnYm2VuPuCXmpuYvmCXQ==",
-            "dev": true,
             "funding": [
                 {
                     "type": "github",

package.json

@@ -37,6 +37,7 @@
         "@ai-sdk/deepseek": "^2.0.0",
         "@ai-sdk/gateway": "^3.0.0",
         "@ai-sdk/google": "^3.0.0",
+        "@ai-sdk/google-vertex": "^4.0.16",
         "@ai-sdk/openai": "^3.0.0",
         "@ai-sdk/react": "^3.0.1",
         "@aws-sdk/client-dynamodb": "^3.957.0",
@@ -49,6 +50,7 @@
         "@next/third-parties": "^16.0.6",
         "@opennextjs/cloudflare": "1.14.8",
         "@openrouter/ai-sdk-provider": "^1.5.4",
+        "@opentelemetry/api": "^1.9.0",
         "@opentelemetry/exporter-trace-otlp-http": "^0.209.0",
         "@opentelemetry/sdk-trace-node": "^2.2.0",
         "@radix-ui/react-alert-dialog": "^1.1.15",

packages/mcp-server/package.json

@@ -1,6 +1,6 @@
 {
     "name": "@next-ai-drawio/mcp-server",
-    "version": "0.1.12",
+    "version": "0.1.13",
     "description": "MCP server for Next AI Draw.io - AI-powered diagram generation with real-time browser preview",
     "type": "module",
     "main": "dist/index.js",

packages/mcp-server/src/index.ts

@@ -260,6 +260,7 @@ COMMON STYLES:
             // Update session state
             currentSession.xml = xml
             currentSession.version++
+            currentSession.lastGetDiagramTime = Date.now()
 
             // Push to embedded server state
             setState(currentSession.id, xml)

tests/unit/server-model-config.test.ts

@@ -0,0 +1,85 @@
+import { afterEach, describe, expect, it } from "vitest"
+import {
+    loadFlattenedServerModels,
+    type ServerModelsConfig,
+    ServerModelsConfigSchema,
+} from "@/lib/server-model-config"
+
+const ORIGINAL_ENV = { ...process.env }
+
+afterEach(() => {
+    process.env.AI_PROVIDER = ORIGINAL_ENV.AI_PROVIDER
+    process.env.AI_MODEL = ORIGINAL_ENV.AI_MODEL
+    process.env.AI_MODELS_CONFIG_PATH = ORIGINAL_ENV.AI_MODELS_CONFIG_PATH
+    process.env.AI_MODELS_CONFIG = ORIGINAL_ENV.AI_MODELS_CONFIG
+})
+
+describe("ServerModelsConfigSchema", () => {
+    it("accepts valid provider names", () => {
+        const config: ServerModelsConfig = {
+            providers: [
+                {
+                    name: "OpenAI Server",
+                    provider: "openai",
+                    models: ["gpt-4o"],
+                },
+            ],
+        }
+
+        expect(() => ServerModelsConfigSchema.parse(config)).not.toThrow()
+    })
+
+    it("rejects invalid provider names", () => {
+        const invalidConfig = {
+            providers: [
+                {
+                    name: "Invalid Provider",
+                    // Cast to any so we can verify runtime validation, not TypeScript
+                    provider: "invalid-provider" as any,
+                    models: ["model-1"],
+                },
+            ],
+        }
+
+        expect(() =>
+            ServerModelsConfigSchema.parse(invalidConfig as any),
+        ).toThrow()
+    })
+})
+
+describe("loadFlattenedServerModels", () => {
+    it("returns empty array when config file is missing", async () => {
+        // Point to a non-existent config path so fs.readFile throws ENOENT
+        process.env.AI_MODELS_CONFIG_PATH = `non-existent-config-${Date.now()}.json`
+
+        const models = await loadFlattenedServerModels()
+        expect(models).toEqual([])
+    })
+
+    it("flattens providers and marks default model from env var config", async () => {
+        // Use AI_MODELS_CONFIG env var instead of file
+        const config: ServerModelsConfig = {
+            providers: [
+                {
+                    name: "OpenAI Server",
+                    provider: "openai",
+                    models: ["gpt-4o", "gpt-4o-mini"],
+                    default: true,
+                },
+            ],
+        }
+        process.env.AI_MODELS_CONFIG = JSON.stringify(config)
+        process.env.AI_MODELS_CONFIG_PATH = "" // Clear file path
+
+        const models = await loadFlattenedServerModels()
+
+        expect(models.length).toBe(2)
+
+        const defaults = models.filter((m) => m.isDefault)
+        expect(defaults.length).toBe(1)
+
+        const defaultModel = defaults[0]
+        expect(defaultModel.provider).toBe("openai")
+        expect(defaultModel.modelId).toBe("gpt-4o") // First model of default provider
+    })
+})