fix(server): count quota by the key actually used, and more review fixes

Found by the PR review, each with a test that failed first:
- Quota: any key header skipped it, even one the provider never reads
  (x-aws-access-key-id with OpenAI), so a request ran on the server's
  key without being counted. The check now runs after the model is
  resolved and uses usesServerCredentials. On main already.
- usesServerCredentials read the raw base URL; "/" cleans up to none, so
  an Ollama request ran on the server's key past the server-model check.
- SGLang's default 127.0.0.1:8000 only fills the settings form. Chat and
  the model list used it as a real address, so the server called its own
  machine even with private URLs blocked. Now a base URL is required.
- With a user's OpenAI key and no base URL, the SDK read the server's
  OPENAI_BASE_URL. The official endpoint is now passed. On main already.
- The Test button refused nothing on the server's keys (Ollama Cloud),
  and a 15 s timeout reported "connected, no tool call".
- The model list for Ollama without a base URL came from ollama.com while
  chat went to the server's Ollama.
- Bedrock's "Too many tokens, please wait" counted as context too long.
- On the server's keys the provider's error text stays in the server log;
  it can name the server's AWS account, role or internal hosts.
- Desktop app: the preset keys are the user's own (NEXT_AI_DRAWIO_DESKTOP),
  so Max Output Tokens can be raised and keyless models in settings work
  again. A launch that found the remembered port taken no longer replaces
  it, which hid the user's chats and settings for good.

app/api/chat/route.ts

@@ -133,36 +133,6 @@ async function handleChatRequest(req: Request): Promise<Response> {
         userId: userId,
     })
 
-    // === SERVER-SIDE QUOTA CHECK START ===
-    // Quota is opt-in: only enabled when DYNAMODB_QUOTA_TABLE env var is set
-    const hasOwnApiKey = !!(
-        req.headers.get("x-ai-provider") &&
-        (req.headers.get("x-ai-api-key") ||
-            req.headers.get("x-aws-access-key-id") ||
-            req.headers.get("x-vertex-api-key"))
-    )
-
-    // Skip quota check if: quota disabled, user has own API key, or is anonymous
-    if (isQuotaEnabled() && !hasOwnApiKey && userId !== "anonymous") {
-        const quotaCheck = await checkAndIncrementRequest(userId, {
-            requests: Number(process.env.DAILY_REQUEST_LIMIT) || 10,
-            tokens: Number(process.env.DAILY_TOKEN_LIMIT) || 200000,
-            tpm: Number(process.env.TPM_LIMIT) || 20000,
-        })
-        if (!quotaCheck.allowed) {
-            return Response.json(
-                {
-                    error: quotaCheck.error,
-                    type: quotaCheck.type,
-                    used: quotaCheck.used,
-                    limit: quotaCheck.limit,
-                },
-                { status: 429 },
-            )
-        }
-    }
-    // === SERVER-SIDE QUOTA CHECK END ===
-
     // === FILE VALIDATION START ===
     const fileValidation = validateFileParts(messages)
     if (!fileValidation.valid) {
@@ -292,14 +262,41 @@ async function handleChatRequest(req: Request): Promise<Response> {
         )
     }
 
+    // === SERVER-SIDE QUOTA CHECK START ===
+    // Quota is opt-in (DYNAMODB_QUOTA_TABLE) and counts what runs on the
+    // server's keys. Decided by the key actually used: a key header the
+    // provider never reads must not skip it.
+    const countsQuota =
+        isQuotaEnabled() && onServerCredentials && userId !== "anonymous"
+    if (countsQuota) {
+        const quotaCheck = await checkAndIncrementRequest(userId, {
+            requests: Number(process.env.DAILY_REQUEST_LIMIT) || 10,
+            tokens: Number(process.env.DAILY_TOKEN_LIMIT) || 200000,
+            tpm: Number(process.env.TPM_LIMIT) || 20000,
+        })
+        if (!quotaCheck.allowed) {
+            return Response.json(
+                {
+                    error: quotaCheck.error,
+                    type: quotaCheck.type,
+                    used: quotaCheck.used,
+                    limit: quotaCheck.limit,
+                },
+                { status: 429 },
+            )
+        }
+    }
+    // === SERVER-SIDE QUOTA CHECK END ===
+
     // Retry once if the provider rejects the requested budget, or (newer
     // Claude models) the sampling or thinking settings
     const model = withOutputTokenLimitFallback(
         withDeprecatedParamsFallback(baseModel),
     )
 
-    // The user setting can raise the budget only on their own key (desktop users
-    // can still raise it themselves); on the server's keys it can only lower it
+    // The user setting can raise the budget only on their own key (in the
+    // desktop app every key is the user's); on the server's keys it can only
+    // lower it
     const maxOutputTokens = resolveMaxOutputTokens(
         req.headers.get("x-max-output-tokens"),
         onServerCredentials,
@@ -587,12 +584,7 @@ IMPORTANT: The "Current diagram XML" is the SINGLE SOURCE OF TRUTH for what's on
             // Record token usage for server-side quota tracking (if enabled)
             // Use totalUsage (cumulative across all steps) instead of usage (final step only)
             // inputTokens already includes cache reads and writes in AI SDK 6
-            if (
-                isQuotaEnabled() &&
-                !hasOwnApiKey &&
-                userId !== "anonymous" &&
-                totalUsage
-            ) {
+            if (countsQuota && totalUsage) {
                 const totalTokens =
                     (totalUsage.inputTokens || 0) +
                     (totalUsage.outputTokens || 0)
@@ -724,7 +716,7 @@ Call this tool to get shape names and usage syntax for a specific library.`,
 
     const response = result.toUIMessageStreamResponse({
         sendReasoning: true,
-        onError: streamErrorText,
+        onError: (error) => streamErrorText(error, onServerCredentials),
         messageMetadata: ({ part }) => {
             if (part.type === "finish") {
                 const usage = (part as any).totalUsage

app/api/validate-model/route.ts

@@ -2,14 +2,15 @@ import { streamText, tool } from "ai"
 import { NextResponse } from "next/server"
 import { z } from "zod"
 import { checkAccessCode } from "@/lib/access-code"
-import { getAIModel } from "@/lib/ai-providers"
+import { getAIModel, usesServerCredentials } from "@/lib/ai-providers"
 import { classifyLLMError } from "@/lib/llm-errors"
 import { allowPrivateUrls, isPrivateUrl } from "@/lib/ssrf-protection"
+import type { ProviderName } from "@/lib/types/model-config"
 
 export const runtime = "nodejs"
 
 interface ValidateRequest {
-    provider: string
+    provider: ProviderName
     apiKey: string
     baseUrl?: string
     modelId: string
@@ -93,6 +94,22 @@ export async function POST(req: Request) {
                 { status: 400 },
             )
         }
+        // The Test button checks the user's own provider. On the server's
+        // keys (Ollama Cloud without a key or URL) anyone could run any model.
+        if (
+            usesServerCredentials(provider, {
+                apiKey,
+                baseUrl,
+                awsAccessKeyId,
+                awsSecretAccessKey,
+                vertexApiKey,
+            })
+        ) {
+            return NextResponse.json(
+                { valid: false, error: "API key is required" },
+                { status: 400 },
+            )
+        }
 
         // The same model the chat would use. A client base URL makes it
         // refuse redirects to internal hosts.
@@ -130,6 +147,14 @@ export async function POST(req: Request) {
         let finishReason: string | undefined
         for await (const part of result.fullStream) {
             if (part.type === "error") throw part.error
+            // The timeout ends the stream with an abort part, not an error
+            if (part.type === "abort") {
+                const timeout = new Error(
+                    `The model did not answer within ${TEST_TIMEOUT_MS / 1000} s.`,
+                )
+                timeout.name = "TimeoutError"
+                throw timeout
+            }
             if (part.type === "tool-call") {
                 calledTool = true
                 break