diff --git a/packages/ai/CHANGELOG.md b/packages/ai/CHANGELOG.md
index 35bb6a355..0167d8ab1 100644
--- a/packages/ai/CHANGELOG.md
+++ b/packages/ai/CHANGELOG.md
@@ -1,6 +1,12 @@
 # Changelog
 
 ## [Unreleased]
+
+### Changed
+
+- Implemented time-budget-based retry strategy for Google Gemini rate limits (429 errors), allowing retries within a 5-minute window instead of a fixed attempt cap
+- Set default `maxRetries` to 5 for Anthropic, Azure OpenAI, and OpenAI client configurations to improve resilience against transient failures
+
 ### Fixed
 
 - Improved error messages for OAuth token refresh failures by including detailed error information from the provider
diff --git a/packages/ai/src/providers/anthropic.ts b/packages/ai/src/providers/anthropic.ts
index 9e77e1666..03ec33238 100644
--- a/packages/ai/src/providers/anthropic.ts
+++ b/packages/ai/src/providers/anthropic.ts
@@ -629,6 +629,7 @@ function createClient(
 		baseURL: config.baseURL,
 		defaultHeaders: config.defaultHeaders,
 		dangerouslyAllowBrowser: config.dangerouslyAllowBrowser,
+		maxRetries: 5,
 	});
 
 	return { client, isOAuthToken: config.isOAuthToken };
diff --git a/packages/ai/src/providers/azure-openai-responses.ts b/packages/ai/src/providers/azure-openai-responses.ts
index 486a49cb5..2075b689d 100644
--- a/packages/ai/src/providers/azure-openai-responses.ts
+++ b/packages/ai/src/providers/azure-openai-responses.ts
@@ -424,6 +424,7 @@ function createClient(model: Model<"azure-openai-responses">, apiKey: string, op
 		apiKey,
 		apiVersion,
 		dangerouslyAllowBrowser: true,
+		maxRetries: 5,
 		defaultHeaders: headers,
 		baseURL: baseUrl,
 	});
diff --git a/packages/ai/src/providers/google-gemini-cli.ts b/packages/ai/src/providers/google-gemini-cli.ts
index 481da9688..8cf459ed9 100644
--- a/packages/ai/src/providers/google-gemini-cli.ts
+++ b/packages/ai/src/providers/google-gemini-cli.ts
@@ -99,6 +99,7 @@ const MAX_RETRIES = 3;
 const BASE_DELAY_MS = 1000;
 const MAX_EMPTY_STREAM_RETRIES = 2;
 const EMPTY_STREAM_BASE_DELAY_MS = 500;
+const RATE_LIMIT_BUDGET_MS = 5 * 60 * 1000;
 const CLAUDE_THINKING_BETA_HEADER = "interleaved-thinking-2025-05-14";
 
 /**
@@ -360,8 +361,9 @@ export const streamGoogleGeminiCli: StreamFunction<"google-gemini-cli"> = (
 			let response: Response | undefined;
 			let lastError: Error | undefined;
 			let requestUrl: string | undefined;
+			let rateLimitTimeSpent = 0;
 
-			for (let attempt = 0; attempt <= MAX_RETRIES; attempt++) {
+			for (let attempt = 0; ; attempt++) {
 				if (options?.signal?.aborted) {
 					throw new Error("Request was aborted");
 				}
@@ -382,13 +384,25 @@ export const streamGoogleGeminiCli: StreamFunction<"google-gemini-cli"> = (
 
 					const errorText = await response.text();
 
-					// Check if retryable
-					if (attempt < MAX_RETRIES && isRetryableError(response.status, errorText)) {
-						// Use server-provided delay or exponential backoff
+					// Handle 429 rate limits with time budget
+					if (response.status === 429) {
+						const serverDelay = extractRetryDelay(errorText, response);
+						if (serverDelay && rateLimitTimeSpent + serverDelay <= RATE_LIMIT_BUDGET_MS) {
+							rateLimitTimeSpent += serverDelay;
+							await abortableSleep(serverDelay, options?.signal);
+							continue;
+						}
+						// Fallback: use exponential backoff if no server delay, up to MAX_RETRIES
+						if (!serverDelay && attempt < MAX_RETRIES) {
+							await abortableSleep(BASE_DELAY_MS * 2 ** attempt, options?.signal);
+							continue;
+						}
+					} else if (attempt < MAX_RETRIES && isRetryableError(response.status, errorText)) {
+						// Non-429 retryable errors use standard attempt cap
 						const serverDelay = extractRetryDelay(errorText, response);
 						const delayMs = serverDelay ?? BASE_DELAY_MS * 2 ** attempt;
 
-						// Check if server delay exceeds max allowed (default: 60s)
+						// Check if server delay exceeds max allowed (default: 60s) for non-429 errors
 						const maxDelayMs = options?.maxRetryDelayMs ?? 60000;
 						if (maxDelayMs > 0 && serverDelay && serverDelay > maxDelayMs) {
 							const delaySeconds = Math.ceil(serverDelay / 1000);
@@ -401,7 +415,7 @@ export const streamGoogleGeminiCli: StreamFunction<"google-gemini-cli"> = (
 						continue;
 					}
 
-					// Not retryable or max retries exceeded
+					// Not retryable or budget exceeded
 					throw new Error(`Cloud Code Assist API error (${response.status}): ${extractErrorMessage(errorText)}`);
 				} catch (error) {
 					// Check for abort - fetch throws AbortError, our code throws "Request was aborted"
diff --git a/packages/ai/src/providers/openai-completions.ts b/packages/ai/src/providers/openai-completions.ts
index 810f64b73..cc94c5bb3 100644
--- a/packages/ai/src/providers/openai-completions.ts
+++ b/packages/ai/src/providers/openai-completions.ts
@@ -480,6 +480,7 @@ async function createClient(
 		apiKey,
 		baseURL: model.baseUrl,
 		dangerouslyAllowBrowser: true,
+		maxRetries: 5,
 		defaultHeaders: headers,
 	});
 }
diff --git a/packages/ai/src/providers/openai-responses.ts b/packages/ai/src/providers/openai-responses.ts
index 833e9236a..20c2f9054 100644
--- a/packages/ai/src/providers/openai-responses.ts
+++ b/packages/ai/src/providers/openai-responses.ts
@@ -397,6 +397,7 @@ function createClient(
 		apiKey,
 		baseURL: model.baseUrl,
 		dangerouslyAllowBrowser: true,
+		maxRetries: 5,
 		defaultHeaders: headers,
 	});
 }