Backport: fix(provider/google): fix Gemini service tier enum after upstream update (#14091)

vercel-ai-sdk[bot] · felixarntz · web-flow · commit 0f2b2f11f272 · 2026-04-03T01:13:05.000Z
This is an automated backport of #14090 to the release-v6.0 branch. FYI @felixarntz Co-authored-by: Felix Arntz <felix.arntz@vercel.com>
diff --git a/.changeset/cold-avocados-tap.md b/.changeset/cold-avocados-tap.md
@@ -0,0 +1,5 @@
+---
+"@ai-sdk/google": patch
+---
+
+fix(provider/google): fix Gemini service tier enum after upstream update
diff --git a/content/providers/01-ai-sdk-providers/15-google-generative-ai.mdx b/content/providers/01-ai-sdk-providers/15-google-generative-ai.mdx
@@ -245,11 +245,11 @@ The following optional provider options are available for Google Generative AI m
   Optional. Defines labels used in billing reports. Available on Vertex AI only.
   See [Google Cloud labels documentation](https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/add-labels-to-api-calls).
 
-- **serviceTier** _'SERVICE_TIER_STANDARD' | 'SERVICE_TIER_FLEX' | 'SERVICE_TIER_PRIORITY'_
+- **serviceTier** _'standard' | 'flex' | 'priority'_
 
   Optional. The service tier to use for the request.
-  Set to 'SERVICE_TIER_FLEX' for 50% cheaper processing at the cost of increased latency.
-  Set to 'SERVICE_TIER_PRIORITY' for ultra-low latency at a 75-100% price premium over 'SERVICE_TIER_STANDARD'.
+  Set to 'flex' for 50% cheaper processing at the cost of increased latency.
+  Set to 'priority' for ultra-low latency at a 75-100% price premium over 'standard'.
 
 - **threshold** _string_
 
diff --git a/examples/ai-functions/src/generate-text/google/service-tier.ts b/examples/ai-functions/src/generate-text/google/service-tier.ts
@@ -8,7 +8,7 @@ run(async () => {
     prompt: 'What color is the sky in one word?',
     providerOptions: {
       google: {
-        serviceTier: 'SERVICE_TIER_FLEX',
+        serviceTier: 'flex',
       } satisfies GoogleLanguageModelOptions,
     },
   });
diff --git a/examples/ai-functions/src/stream-text/google/service-tier.ts b/examples/ai-functions/src/stream-text/google/service-tier.ts
@@ -8,7 +8,7 @@ run(async () => {
     prompt: 'What color is the sky in one word?',
     providerOptions: {
       google: {
-        serviceTier: 'SERVICE_TIER_FLEX',
+        serviceTier: 'priority',
       } satisfies GoogleLanguageModelOptions,
     },
   });
diff --git a/packages/google/src/google-generative-ai-language-model.test.ts b/packages/google/src/google-generative-ai-language-model.test.ts
@@ -566,13 +566,13 @@ describe('doGenerate', () => {
       prompt: TEST_PROMPT,
       providerOptions: {
         google: {
-          serviceTier: 'SERVICE_TIER_FLEX',
+          serviceTier: 'flex',
         },
       },
     });
 
     expect(await server.calls[0].requestBodyJson).toMatchObject({
-      serviceTier: 'SERVICE_TIER_FLEX',
+      serviceTier: 'flex',
     });
   });
 
@@ -606,15 +606,15 @@ describe('doGenerate', () => {
           candidatesTokenCount: 2,
           totalTokenCount: 3,
         },
-        serviceTier: 'SERVICE_TIER_FLEX',
+        serviceTier: 'flex',
       },
     };
 
     const { providerMetadata } = await model.doGenerate({
       prompt: TEST_PROMPT,
     });
 
-    expect(providerMetadata?.google.serviceTier).toBe('SERVICE_TIER_FLEX');
+    expect(providerMetadata?.google.serviceTier).toBe('flex');
   });
 
   it('should expose null serviceTier in provider metadata when not present', async () => {
@@ -4014,7 +4014,7 @@ describe('doStream', () => {
             candidatesTokenCount: 2,
             totalTokenCount: 3,
           },
-          serviceTier: 'SERVICE_TIER_FLEX',
+          serviceTier: 'flex',
         })}\n\n`,
       ],
     };
@@ -4029,7 +4029,7 @@ describe('doStream', () => {
     expect(
       finishEvent?.type === 'finish' &&
         finishEvent.providerMetadata?.google.serviceTier,
-    ).toBe('SERVICE_TIER_FLEX');
+    ).toBe('flex');
   });
 
   it('should expose null serviceTier in provider metadata on finish when not present', async () => {
diff --git a/packages/google/src/google-generative-ai-options.ts b/packages/google/src/google-generative-ai-options.ts
@@ -192,13 +192,7 @@ export const googleLanguageModelOptions = lazySchema(() =>
       /**
        * Optional. The service tier to use for the request.
        */
-      serviceTier: z
-        .enum([
-          'SERVICE_TIER_STANDARD',
-          'SERVICE_TIER_FLEX',
-          'SERVICE_TIER_PRIORITY',
-        ])
-        .optional(),
+      serviceTier: z.enum(['standard', 'flex', 'priority']).optional(),
     }),
   ),
 );

-Original file line number
+Diff line change
@@ @@ -0,0 +1,5 @@ @@
 +---
 +"@ai-sdk/google": patch
 +---
++
 +fix(provider/google): fix Gemini service tier enum after upstream update