Skip to content

An AI-powered search engine with a generative UI

feat: Add quality mode rate limiting with user-friendly error handling

Yoshiki Miuraa0891caecbde
diff --git a/.env.local.example b/.env.local.example
index ec5f4710..bce78d30 100644
--- a/.env.local.example
+++ b/.env.local.example
@@ -75,6 +75,14 @@ NEXT_PUBLIC_SUPABASE_ANON_KEY=[YOUR_SUPABASE_ANON_KEY]
# Enable additional functionality as needed
# =============================================================================
+# -----------------------------------------------------------------------------
+# Cloud Deployment Configuration
+# -----------------------------------------------------------------------------
+# Enable rate limiting for cloud deployments
+# When enabled, quality mode chat requests are limited to 5 per day
+# Requires UPSTASH_REDIS_REST_URL and UPSTASH_REDIS_REST_TOKEN to be configured
+# MORPHIC_CLOUD_DEPLOYMENT=true
+
# -----------------------------------------------------------------------------
# File Upload Support (Cloudflare R2)
# Enable file and image upload functionality
@@ -130,8 +138,5 @@ NEXT_PUBLIC_SUPABASE_ANON_KEY=[YOUR_SUPABASE_ANON_KEY]
# -----------------------------------------------------------------------------
# Advanced Configuration
# -----------------------------------------------------------------------------
-# Model Configuration Profile
-# MORPHIC_MODELS_PROFILE=default # Options: default, cloud, or custom profile
-
# Feedback Notifications (Slack)
# SLACK_WEBHOOK_URL=[YOUR_SLACK_WEBHOOK_URL]
diff --git a/app/api/chat/route.ts b/app/api/chat/route.ts
index 93d1e17c..67f0f473 100644
--- a/app/api/chat/route.ts
+++ b/app/api/chat/route.ts
@@ -2,6 +2,7 @@ import { revalidateTag } from 'next/cache'
import { cookies } from 'next/headers'
import { getCurrentUserId } from '@/lib/auth/get-current-user'
+import { checkAndEnforceQualityLimit } from '@/lib/rate-limit/quality-limit'
import { createChatStreamResponse } from '@/lib/streaming/create-chat-stream-response'
import { SearchMode } from '@/lib/types/search'
import { selectModel } from '@/lib/utils/model-selection'
@@ -93,6 +94,14 @@ export async function POST(req: Request) {
)
}
+ // Check rate limit for quality mode
+ const modelTypeCookie = cookieStore.get('modelType')?.value
+ const rateLimitResponse = await checkAndEnforceQualityLimit(
+ userId,
+ modelTypeCookie === 'quality'
+ )
+ if (rateLimitResponse) return rateLimitResponse
+
const streamStart = performance.now()
perfLog(
`createChatStreamResponse - Start: model=${selectedModel.providerId}:${selectedModel.id}, searchMode=${searchMode}`
diff --git a/bun.lock b/bun.lock
index a7dbf413..8442f041 100644
--- a/bun.lock
+++ b/bun.lock
@@ -36,7 +36,7 @@
"@supabase/ssr": "^0.6.1",
"@supabase/supabase-js": "^2.49.4",
"@tailwindcss/typography": "^0.5.12",
- "@upstash/redis": "^1.34.0",
+ "@upstash/redis": "^1.35.5",
"@vercel/analytics": "^1.5.0",
"@vercel/otel": "^1.13.0",
"ai": "^5.0.39",
@@ -964,7 +964,7 @@
"@unrs/resolver-binding-win32-x64-msvc": ["@unrs/[email protected]", "", { "os": "win32", "cpu": "x64" }, "sha512-lrW200hZdbfRtztbygyaq/6jP6AKE8qQN2KvPcJ+x7wiD038YtnYtZ82IMNJ69GJibV7bwL3y9FgK+5w/pYt6g=="],
- "@upstash/redis": ["@upstash/[email protected]", "", { "dependencies": { "uncrypto": "^0.1.3" } }, "sha512-hSjv66NOuahW3MisRGlSgoszU2uONAY2l5Qo3Sae8OT3/Tng9K+2/cBRuyPBX8egwEGcNNCF9+r0V6grNnhL+w=="],
+ "@upstash/redis": ["@upstash/[email protected]", "", { "dependencies": { "uncrypto": "^0.1.3" } }, "sha512-KdLdNAspQGOTGeC++o2LDBzNbMXrfInnmW5nUJfNXabnVh8X4NPrlJ0X4j75cBUShiMpXB3uI1ql4KpFQeqrHQ=="],
"@vercel/analytics": ["@vercel/[email protected]", "", { "peerDependencies": { "@remix-run/react": "^2", "@sveltejs/kit": "^1 || ^2", "next": ">= 13", "react": "^18 || ^19 || ^19.0.0-rc", "svelte": ">= 4", "vue": "^3", "vue-router": "^4" }, "optionalPeers": ["@remix-run/react", "@sveltejs/kit", "next", "react", "svelte", "vue", "vue-router"] }, "sha512-MYsBzfPki4gthY5HnYN7jgInhAZ7Ac1cYDoRWFomwGHWEX7odTEzbtg9kf/QSo7XEsEAqlQugA6gJ2WS2DEa3g=="],
diff --git a/components/chat-error.tsx b/components/chat-error.tsx
index 838c2182..0cf3ac63 100644
--- a/components/chat-error.tsx
+++ b/components/chat-error.tsx
@@ -9,7 +9,20 @@ interface ChatErrorProps {
export function ChatError({ error }: ChatErrorProps) {
if (!error) return null
- const errorMessage = error instanceof Error ? error.message : error
+ let errorMessage = error instanceof Error ? error.message : error
+
+ // Try to parse JSON error response and extract user-friendly message
+ try {
+ const jsonMatch = errorMessage?.match(/\{.*\}/)
+ if (jsonMatch) {
+ const parsed = JSON.parse(jsonMatch[0])
+ if (parsed.error) {
+ errorMessage = parsed.error
+ }
+ }
+ } catch {
+ // If parsing fails, use the original error message
+ }
return (
<Card className="border-destructive bg-destructive/10 p-4">
diff --git a/components/chat.tsx b/components/chat.tsx
index 1a7c0907..2685ce59 100644
--- a/components/chat.tsx
+++ b/components/chat.tsx
@@ -110,13 +110,29 @@ export function Chat({
const isRateLimit =
error.message?.includes('429') ||
errorMessage.includes('rate limit') ||
- errorMessage.includes('too many requests')
+ errorMessage.includes('too many requests') ||
+ errorMessage.includes('daily limit')
if (isRateLimit) {
+ // Try to parse JSON error response for quality mode rate limit
+ let parsedError: { error?: string; resetAt?: number; remaining?: number } = {}
+ try {
+ // Extract JSON from error message if it exists
+ const jsonMatch = error.message?.match(/\{.*\}/)
+ if (jsonMatch) {
+ parsedError = JSON.parse(jsonMatch[0])
+ }
+ } catch {
+ // Ignore parse errors
+ }
+
+ // Use parsed error message or fallback
+ const userMessage = parsedError.error || 'You have reached your daily limit for quality mode chat requests.'
+
setErrorModal({
open: true,
type: 'rate-limit',
- message: error.message,
+ message: userMessage,
details: undefined
})
} else if (
diff --git a/components/error-modal.tsx b/components/error-modal.tsx
index 1c628735..38227e5c 100644
--- a/components/error-modal.tsx
+++ b/components/error-modal.tsx
@@ -57,7 +57,7 @@ export function ErrorModal({
const getErrorDescription = () => {
switch (error.type) {
case 'rate-limit':
- return 'You have made too many requests. Please wait a moment before trying again.'
+ return error.message || 'You have made too many requests. Please wait a moment before trying again.'
case 'auth':
return 'You need to sign in to continue using this feature.'
case 'forbidden':
@@ -71,7 +71,7 @@ export function ErrorModal({
const getErrorDetails = () => {
if (error.type === 'rate-limit') {
- return 'Our rate limiting helps ensure fair usage and maintain service quality for all users. The limit will reset shortly.'
+ return 'The limit will reset at midnight UTC. You can continue using speed mode without restrictions.'
}
return error.details
}
diff --git a/lib/config/load-models-config.ts b/lib/config/load-models-config.ts
index 4a4080f9..fca0f0aa 100644
--- a/lib/config/load-models-config.ts
+++ b/lib/config/load-models-config.ts
@@ -21,7 +21,8 @@ const VALID_MODEL_TYPES: ModelType[] = ['speed', 'quality']
const VALID_SEARCH_MODES: SearchMode[] = ['quick', 'adaptive', 'planning']
function resolveConfigPath(): string {
- const profile = process.env.MORPHIC_MODELS_PROFILE?.trim() || 'default'
+ const profile =
+ process.env.MORPHIC_CLOUD_DEPLOYMENT === 'true' ? 'cloud' : 'default'
const file = `${profile}.json`
const configPath = path.resolve(process.cwd(), 'config', 'models', file)
return configPath
@@ -68,7 +69,8 @@ function validateModelsConfigStructure(
}
export async function loadModelsConfig(): Promise<ModelsConfig> {
- const profile = process.env.MORPHIC_MODELS_PROFILE?.trim() || 'default'
+ const profile =
+ process.env.MORPHIC_CLOUD_DEPLOYMENT === 'true' ? 'cloud' : 'default'
if (cachedConfig && cachedProfile === profile) {
return cachedConfig
@@ -106,7 +108,8 @@ export async function loadModelsConfig(): Promise<ModelsConfig> {
// Synchronous load (for code paths that need sync access)
export function loadModelsConfigSync(): ModelsConfig {
- const profile = process.env.MORPHIC_MODELS_PROFILE?.trim() || 'default'
+ const profile =
+ process.env.MORPHIC_CLOUD_DEPLOYMENT === 'true' ? 'cloud' : 'default'
if (cachedConfig && cachedProfile === profile) {
return cachedConfig
}
diff --git a/lib/rate-limit/quality-limit.ts b/lib/rate-limit/quality-limit.ts
new file mode 100644
index 00000000..1004eef8
--- /dev/null
+++ b/lib/rate-limit/quality-limit.ts
@@ -0,0 +1,128 @@
+import { Redis } from '@upstash/redis'
+
+import { perfLog } from '@/lib/utils/perf-logging'
+
+const DAILY_QUALITY_LIMIT = 5
+
+/**
+ * Internal function to check rate limit using Upstash Redis
+ */
+async function checkQualityChatLimit(userId: string): Promise<{
+ allowed: boolean
+ remaining: number
+ resetAt: number
+}> {
+ // If not in cloud deployment mode, allow unlimited requests
+ if (process.env.MORPHIC_CLOUD_DEPLOYMENT !== 'true') {
+ return { allowed: true, remaining: Infinity, resetAt: 0 }
+ }
+
+ // If Upstash is not configured, allow unlimited requests
+ if (
+ !process.env.UPSTASH_REDIS_REST_URL ||
+ !process.env.UPSTASH_REDIS_REST_TOKEN
+ ) {
+ return { allowed: true, remaining: Infinity, resetAt: 0 }
+ }
+
+ try {
+ const redis = new Redis({
+ url: process.env.UPSTASH_REDIS_REST_URL,
+ token: process.env.UPSTASH_REDIS_REST_TOKEN
+ })
+
+ // Create daily key with date
+ const dateKey = new Date().toISOString().split('T')[0] // YYYY-MM-DD
+ const key = `rl:quality:chat:${userId}:${dateKey}`
+
+ // Increment counter with timeout
+ const count = await Promise.race([
+ redis.incr(key),
+ new Promise<number>((_, reject) =>
+ setTimeout(() => reject(new Error('Redis timeout')), 3000)
+ )
+ ])
+
+ // Set expiry on first request of the day
+ if (count === 1) {
+ const secondsUntilMidnight = getSecondsUntilMidnight()
+ await redis.expire(key, secondsUntilMidnight)
+ }
+
+ const remaining = Math.max(0, DAILY_QUALITY_LIMIT - count)
+ const resetAt = getNextMidnightTimestamp()
+
+ return {
+ allowed: count <= DAILY_QUALITY_LIMIT,
+ remaining,
+ resetAt
+ }
+ } catch (error) {
+ // On error, allow request to proceed (availability over strict limiting)
+ console.error('Rate limit check failed:', error)
+ return { allowed: true, remaining: Infinity, resetAt: 0 }
+ }
+}
+
+/**
+ * Get seconds until next midnight UTC
+ */
+function getSecondsUntilMidnight(): number {
+ const now = new Date()
+ const midnight = new Date(now)
+ midnight.setUTCHours(24, 0, 0, 0)
+ return Math.floor((midnight.getTime() - now.getTime()) / 1000)
+}
+
+/**
+ * Get timestamp of next midnight UTC
+ */
+function getNextMidnightTimestamp(): number {
+ const now = new Date()
+ const midnight = new Date(now)
+ midnight.setUTCHours(24, 0, 0, 0)
+ return midnight.getTime()
+}
+
+/**
+ * Check and enforce quality chat rate limit
+ * Returns a 429 Response if limit is exceeded, null if allowed
+ */
+export async function checkAndEnforceQualityLimit(
+ userId: string,
+ isQualityMode: boolean
+): Promise<Response | null> {
+ // Only check limit for quality mode
+ if (!isQualityMode) {
+ return null
+ }
+
+ const result = await checkQualityChatLimit(userId)
+
+ if (!result.allowed) {
+ return new Response(
+ JSON.stringify({
+ error: `You've reached your daily limit of ${DAILY_QUALITY_LIMIT} quality mode requests.`,
+ remaining: 0,
+ resetAt: result.resetAt,
+ limit: DAILY_QUALITY_LIMIT
+ }),
+ {
+ status: 429,
+ headers: {
+ 'Content-Type': 'application/json',
+ 'X-RateLimit-Limit': String(DAILY_QUALITY_LIMIT),
+ 'X-RateLimit-Remaining': '0',
+ 'X-RateLimit-Reset': String(result.resetAt)
+ }
+ }
+ )
+ }
+
+ // Log usage for debugging
+ perfLog(
+ `Quality chat usage: ${DAILY_QUALITY_LIMIT - result.remaining}/${DAILY_QUALITY_LIMIT}`
+ )
+
+ return null
+}
diff --git a/package.json b/package.json
index c5e900dc..f267af0b 100644
--- a/package.json
+++ b/package.json
@@ -49,7 +49,7 @@
"@supabase/ssr": "^0.6.1",
"@supabase/supabase-js": "^2.49.4",
"@tailwindcss/typography": "^0.5.12",
- "@upstash/redis": "^1.34.0",
+ "@upstash/redis": "^1.35.5",
"@vercel/analytics": "^1.5.0",
"@vercel/otel": "^1.13.0",
"ai": "^5.0.39",