mirror of
https://github.com/DayuanJiang/next-ai-draw-io.git
synced 2026-09-08 12:20:20 +08:00
[Feature] Server-side multi-provider/model support (#583)
* [Feature] Server side multi-pvorider/model support * copilot suggesition implemented * feat: improve model selector UI and auto-select default server model - Replace emoji headers with Lucide icons (Monitor, User) - Fix transition-all to explicit properties per web guidelines - Use CSS padding instead of hardcoded space indentation - Add ModelSelectorSectionHeader component for section headers - Replace Star icon with "default" text label - Style Configure button with muted text color - Auto-select default server model when page loads - Support AI_MODELS_CONFIG env var for cloud deployments - Support custom apiKeyEnv/baseUrlEnv per provider config * docs: update server-side multi-model configuration documentation - Add AI_MODELS_CONFIG env var option for cloud deployments - Document apiKeyEnv and baseUrlEnv fields for custom env var names - Document default field for auto-selecting default model - Remove deprecated version field from examples - Add field reference table for clarity --------- Co-authored-by: dayuan.jiang <jdy.toh@gmail.com>
This commit is contained in:
Biki Kalita
co-authored by
dayuan.jiang
dayuan.jiang
parent
b128c57e94
commit
b23b9179a0
+2
-1
@@ -68,4 +68,5 @@ CLAUDE.md
|
||||
|
||||
# edgeone
|
||||
.edgeone
|
||||
opencode.json
|
||||
opencode.json
|
||||
ai-models.json
|
||||
|
||||
@@ -222,6 +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.
|
||||
|
||||
+31
-1
@@ -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 +213,8 @@ 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
|
||||
@@ -201,6 +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)
|
||||
|
||||
@@ -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,
|
||||
})
|
||||
}
|
||||
@@ -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>
|
||||
)
|
||||
|
||||
@@ -968,6 +968,10 @@ export default function ChatPanel({
|
||||
"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",
|
||||
}),
|
||||
|
||||
+184
-75
@@ -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"
|
||||
@@ -63,7 +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)
|
||||
@@ -91,10 +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(
|
||||
@@ -161,7 +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
|
||||
@@ -198,83 +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 */}
|
||||
@@ -283,7 +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>
|
||||
|
||||
@@ -214,6 +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。
|
||||
|
||||
@@ -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。
|
||||
|
||||
@@ -232,6 +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).
|
||||
|
||||
@@ -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.
|
||||
|
||||
@@ -215,6 +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。
|
||||
|
||||
@@ -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。
|
||||
|
||||
@@ -101,6 +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)
|
||||
|
||||
@@ -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 +381,8 @@ 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
|
||||
} {
|
||||
@@ -327,6 +396,7 @@ export function getSelectedAIConfig(): {
|
||||
awsSecretAccessKey: "",
|
||||
awsRegion: "",
|
||||
awsSessionToken: "",
|
||||
selectedModelId: "",
|
||||
vertexApiKey: "",
|
||||
}
|
||||
|
||||
@@ -350,6 +420,7 @@ export function getSelectedAIConfig(): {
|
||||
awsSecretAccessKey: "",
|
||||
awsRegion: "",
|
||||
awsSessionToken: "",
|
||||
selectedModelId: "",
|
||||
vertexApiKey: "",
|
||||
}
|
||||
}
|
||||
@@ -361,12 +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 }
|
||||
@@ -383,6 +474,7 @@ export function getSelectedAIConfig(): {
|
||||
awsSecretAccessKey: model.awsSecretAccessKey || "",
|
||||
awsRegion: model.awsRegion || "",
|
||||
awsSessionToken: model.awsSessionToken || "",
|
||||
selectedModelId: config.selectedModelId || "",
|
||||
// Vertex AI credentials (Express Mode)
|
||||
vertexApiKey: model.vertexApiKey || "",
|
||||
}
|
||||
|
||||
+108
-26
@@ -34,6 +34,9 @@ export interface ClientOverrides {
|
||||
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
|
||||
@@ -92,6 +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
|
||||
*/
|
||||
@@ -481,9 +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. ` +
|
||||
@@ -621,7 +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}`)
|
||||
@@ -671,11 +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
|
||||
@@ -694,11 +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({
|
||||
@@ -713,12 +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({
|
||||
@@ -756,11 +809,12 @@ export function getAIModel(overrides?: ClientOverrides): ModelConfig {
|
||||
}
|
||||
|
||||
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
|
||||
@@ -794,11 +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,
|
||||
@@ -809,11 +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({
|
||||
@@ -828,11 +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({
|
||||
@@ -844,11 +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({
|
||||
@@ -957,11 +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
|
||||
@@ -993,11 +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()
|
||||
@@ -1022,11 +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({
|
||||
|
||||
@@ -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"
|
||||
}
|
||||
}
|
||||
|
||||
@@ -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"
|
||||
}
|
||||
}
|
||||
|
||||
@@ -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"
|
||||
}
|
||||
}
|
||||
|
||||
@@ -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
|
||||
}
|
||||
@@ -54,7 +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
|
||||
@@ -69,6 +69,13 @@ export interface FlattenedModel {
|
||||
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
|
||||
@@ -307,7 +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[] = []
|
||||
|
||||
@@ -333,6 +340,8 @@ export function flattenModels(config: MultiModelConfig): FlattenedModel[] {
|
||||
vertexApiKey: provider.vertexApiKey,
|
||||
|
||||
validated: model.validated,
|
||||
source: "user",
|
||||
isDefault: false,
|
||||
})
|
||||
}
|
||||
}
|
||||
|
||||
Generated
+87
-13
@@ -15,7 +15,7 @@
|
||||
"@ai-sdk/deepseek": "^2.0.0",
|
||||
"@ai-sdk/gateway": "^3.0.0",
|
||||
"@ai-sdk/google": "^3.0.0",
|
||||
"@ai-sdk/google-vertex": "^4.0.11",
|
||||
"@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",
|
||||
@@ -207,13 +207,13 @@
|
||||
}
|
||||
},
|
||||
"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"
|
||||
@@ -223,15 +223,15 @@
|
||||
}
|
||||
},
|
||||
"node_modules/@ai-sdk/google-vertex": {
|
||||
"version": "4.0.11",
|
||||
"resolved": "https://registry.npmjs.org/@ai-sdk/google-vertex/-/google-vertex-4.0.11.tgz",
|
||||
"integrity": "sha512-G4pdEboGAp9fEdGfImSB5psOnhKudrsm170ZsTJ1L14WtSRS5S+LgMrOedpoDW1xyTsK7Y4FYnyhLBq1hHJdXA==",
|
||||
"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": "3.0.9",
|
||||
"@ai-sdk/google": "3.0.6",
|
||||
"@ai-sdk/provider": "3.0.2",
|
||||
"@ai-sdk/provider-utils": "4.0.4",
|
||||
"@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": {
|
||||
@@ -241,6 +241,80 @@
|
||||
"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"
|
||||
},
|
||||
"peerDependencies": {
|
||||
"zod": "^3.25.76 || ^4.1.8"
|
||||
}
|
||||
},
|
||||
"node_modules/@ai-sdk/openai": {
|
||||
"version": "3.0.7",
|
||||
"resolved": "https://registry.npmjs.org/@ai-sdk/openai/-/openai-3.0.7.tgz",
|
||||
|
||||
+1
-1
@@ -37,7 +37,7 @@
|
||||
"@ai-sdk/deepseek": "^2.0.0",
|
||||
"@ai-sdk/gateway": "^3.0.0",
|
||||
"@ai-sdk/google": "^3.0.0",
|
||||
"@ai-sdk/google-vertex": "^4.0.11",
|
||||
"@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",
|
||||
|
||||
@@ -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
|
||||
})
|
||||
})
|
||||
Reference in New Issue
Block a user