diff --git a/src/hooks/runtime-fallback/constants.ts b/src/hooks/runtime-fallback/constants.ts
index 206762961..43df0f936 100644
--- a/src/hooks/runtime-fallback/constants.ts
+++ b/src/hooks/runtime-fallback/constants.ts
@@ -41,6 +41,12 @@ export const RETRYABLE_ERROR_PATTERNS = [
   /(?:^|\s)429(?:\s|$)/,
   /(?:^|\s)503(?:\s|$)/,
   /(?:^|\s)529(?:\s|$)/,
+  /使用上限/,
+  /频率限制/,
+  /请求过于频繁/,
+  /暂时不可用/,
+  /服务不可用/,
+  /请稍后重试/,
 ]
 
 /**
diff --git a/src/hooks/runtime-fallback/error-classifier.test.ts b/src/hooks/runtime-fallback/error-classifier.test.ts
index c9eef6e06..a7f43210a 100644
--- a/src/hooks/runtime-fallback/error-classifier.test.ts
+++ b/src/hooks/runtime-fallback/error-classifier.test.ts
@@ -59,6 +59,44 @@ describe("runtime-fallback error classifier", () => {
     expect(retryable).toBe(true)
   })
 
+  test("treats localized transient provider messages as retryable", () => {
+    //#given
+    const errors = [
+      { message: "请求过于频繁，请稍后重试" },
+      { message: "服务暂时不可用" },
+      { message: "触发频率限制" },
+    ]
+
+    //#when
+    const retryable = errors.map((error) => isRetryableError(error, [429, 503, 529]))
+
+    //#then
+    expect(retryable).toEqual([true, true, true])
+  })
+
+  test("classifies localized quota exhaustion messages as quota_exceeded", () => {
+    //#given
+    const errors = [
+      { message: "已达到 5 小时的使用上限" },
+      { message: "已达到每日调用限制" },
+      { message: "额度不足" },
+      { message: "账户余额不足" },
+      { message: "免费额度已耗尽" },
+    ]
+
+    //#when
+    const classifications = errors.map((error) => classifyErrorType(error))
+
+    //#then
+    expect(classifications).toEqual([
+      "quota_exceeded",
+      "quota_exceeded",
+      "quota_exceeded",
+      "quota_exceeded",
+      "quota_exceeded",
+    ])
+  })
+
   test("classifies ProviderModelNotFoundError as model_not_found", () => {
     //#given
     const error = {
diff --git a/src/hooks/runtime-fallback/error-classifier.ts b/src/hooks/runtime-fallback/error-classifier.ts
index 0901a4e05..33b17ccf3 100644
--- a/src/hooks/runtime-fallback/error-classifier.ts
+++ b/src/hooks/runtime-fallback/error-classifier.ts
@@ -143,6 +143,11 @@ export function classifyErrorType(error: unknown): string | undefined {
     /payment.?required/i.test(message) ||
     /usage\s+limit/i.test(message) ||
     /credit\s+balance.*too\s+low/i.test(message) ||
+    /使用上限/.test(message) ||
+    /达到.*限制/.test(message) ||
+    /额度.*不足/.test(message) ||
+    /余额.*不足/.test(message) ||
+    /已耗尽/.test(message) ||
     isLocalizedQuotaExhaustionMessage(message)
   ) {
     return "quota_exceeded"
diff --git a/src/shared/model-error-classifier.test.ts b/src/shared/model-error-classifier.test.ts
index 33fd3f4c1..0ab40e8eb 100644
--- a/src/shared/model-error-classifier.test.ts
+++ b/src/shared/model-error-classifier.test.ts
@@ -216,6 +216,21 @@ describe("model-error-classifier", () => {
     expect(result).toBe(true)
   })
 
+  test("treats localized transient provider messages as retryable", () => {
+    //#given
+    const errors = [
+      { message: "请求过于频繁，请稍后重试" },
+      { message: "服务暂时不可用" },
+      { message: "触发频率限制" },
+    ]
+
+    //#when
+    const results = errors.map((error) => shouldRetryError(error))
+
+    //#then
+    expect(results).toEqual([true, true, true])
+  })
+
   test("treats subscription quota message as non-retryable", () => {
     //#given
     const error = { message: "Subscription quota exceeded. You can continue using free models." }
@@ -227,6 +242,22 @@ describe("model-error-classifier", () => {
     expect(result).toBe(false)
   })
 
+  test("treats localized quota exhaustion messages as non-retryable stop errors", () => {
+    //#given
+    const errors = [
+      { message: "已达到 5 小时的使用上限" },
+      { message: "额度不足" },
+      { message: "账户余额不足" },
+      { message: "免费额度已耗尽" },
+    ]
+
+    //#when
+    const results = errors.map((error) => shouldRetryError(error))
+
+    //#then
+    expect(results).toEqual([false, false, false, false])
+  })
+
   test("treats HTTP 429 rate limit message as retryable", () => {
     //#given
     const error = { message: "429 Too Many Requests: rate limit reached" }
diff --git a/src/shared/model-error-classifier.ts b/src/shared/model-error-classifier.ts
index fb3fdf480..787f4825a 100644
--- a/src/shared/model-error-classifier.ts
+++ b/src/shared/model-error-classifier.ts
@@ -67,6 +67,7 @@ const RETRYABLE_MESSAGE_PATTERNS = [
   "balance",
   "temporarily unavailable",
   "try again",
+  "请稍后重试",
   "503",
   "502",
   "504",
@@ -74,6 +75,11 @@ const RETRYABLE_MESSAGE_PATTERNS = [
   "529",
   "selected provider is forbidden",
   "provider is forbidden",
+  // Chinese retryable patterns (Zhipu, etc.)
+  "频率限制",           // "rate limit"
+  "请求过于频繁",       // "too many requests"
+  "暂时不可用",         // "temporarily unavailable"
+  "服务不可用",         // "service unavailable"
 ]
 
 /**
@@ -106,6 +112,10 @@ const STOP_MESSAGE_PATTERNS = [
   "in arrears",
   "fair use policy",
   "recharge and try",
+  "使用上限",
+  "额度不足",
+  "余额不足",
+  "已耗尽",
 ]
 
 const AUTO_RETRY_GATE_PATTERNS = [