Skip to content

updating the models #10096

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
sedanah-m wants to merge 3 commits into
base: main
Choose a base branch
from
Open

updating the models #10096

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
046fee9
updating the models
sedanah-m Jun 16, 2026
8a9c694
yarn formatting fix
sedanah-m Jun 16, 2026
03f8b97
Apply suggestions from code review
sedanah-m Jun 17, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion e2e/smoke-tests/sample-apps/modular.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -313,7 +313,7 @@ function callPerformance(app) {
async function callAI(app) {
console.log('[AI] start');
const ai = getAI(app, { backend: new VertexAIBackend() });
const model = getGenerativeModel(ai, { model: 'gemini-2.5-flash' });
const model = getGenerativeModel(ai, { model: 'gemini-3.5-flash' });
const result = await model.countTokens('abcdefg');
console.log(`[AI] counted tokens: ${result.totalTokens}`);
}
Expand Down
4 changes: 2 additions & 2 deletions e2e/smoke-tests/tests/modular.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -314,8 +314,8 @@ describe('MODULAR', () => {
ai = getAI(app, { backend: new VertexAIBackend() });
});
it('getGenerativeModel()', async () => {
const model = getGenerativeModel(ai, { model: 'gemini-2.5-flash' });
expect(model.model).toMatch(/gemini-2.5-flash$/);
const model = getGenerativeModel(ai, { model: 'gemini-3.5-flash' });
expect(model.model).toMatch(/gemini-3.5-flash$/);
});
});

Expand Down
4 changes: 2 additions & 2 deletions packages/ai/integration/chat.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -97,15 +97,15 @@ describe('Chat Session', function () {
expect(history[3].role).to.equal('model');
expect(history[3].parts[0].text?.toLowerCase()).to.include('rome');

if (model.model.includes('gemini-2.5-flash')) {
if (model.model.includes('gemini-3.5-flash')) {
// Token counts can vary slightly in chat context
expect(response1.usageMetadata!.promptTokenCount).to.not.equal(0);
expect(response1.usageMetadata!.candidatesTokenCount).to.not.equal(0);
expect(response1.usageMetadata!.totalTokenCount).to.not.equal(0);
expect(response2.usageMetadata!.promptTokenCount).to.not.equal(0);
expect(response2.usageMetadata!.candidatesTokenCount).to.not.equal(0);
expect(response2.usageMetadata!.totalTokenCount).to.not.equal(0);
} else if (model.model.includes('gemini-2.0-flash')) {
} else if (model.model.includes('gemini-3.1-flash-lite')) {
expect(response1.usageMetadata).to.not.be.null;
// Token counts can vary slightly in chat context
expect(response1.usageMetadata!.promptTokenCount).to.not.equal(0);
Expand Down
18 changes: 6 additions & 12 deletions packages/ai/integration/constants.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -62,16 +62,13 @@ const backendNames: Map<BackendType, string> = new Map([
]);

const modelNames: readonly string[] = [
'gemini-2.0-flash-001',
'gemini-2.0-flash-lite-001',
'gemini-2.5-flash',
'gemini-2.5-flash-lite',
'gemini-2.5-pro',
'gemini-3-pro-preview'
'gemini-3.1-flash-lite',
'gemini-3.5-flash',
'gemini-3.1-pro-preview'
];

// Used for testing non-AI behavior (e.g. Network requests). Configured to minimize cost.
export const cheapestModel = 'gemini-2.0-flash';
export const cheapestModel = 'gemini-3.1-flash-lite';
export const defaultAIInstance = getAI(app, { backend: new VertexAIBackend() });
export const defaultGenerativeModel = getGenerativeModel(defaultAIInstance, {
model: cheapestModel,
Expand All @@ -82,11 +79,8 @@ export const defaultGenerativeModel = getGenerativeModel(defaultAIInstance, {

// The Live API requires a different set of models, and they're different for each backend.
const liveModelNames: Map<BackendType, string[]> = new Map([
[BackendType.GOOGLE_AI, ['gemini-2.5-flash-native-audio-preview-09-2025']],
[
BackendType.VERTEX_AI,
['gemini-live-2.5-flash-preview-native-audio-09-2025']
]
[BackendType.GOOGLE_AI, ['gemini-3.1-flash-live-preview']],
[BackendType.VERTEX_AI, ['gemini-3.1-flash-live-preview']]
]);

/**
Expand Down
4 changes: 2 additions & 2 deletions packages/ai/integration/generate-content.test.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -98,7 +98,7 @@ describe('Generate Content', function () {

expect(response.usageMetadata).to.not.be.null;

if (model.model.includes('gemini-2.5-flash')) {
if (model.model.includes('gemini-3.5-flash')) {
expect(response.usageMetadata!.promptTokenCount).to.not.equal(0);
expect(response.usageMetadata!.candidatesTokenCount).to.not.equal(0);
expect(response.usageMetadata!.thoughtsTokenCount).to.not.equal(0);
Expand All @@ -115,7 +115,7 @@ describe('Generate Content', function () {
).to.not.equal(0);

// candidatesTokenDetails comes back about half the time, so let's just not test it.
} else if (model.model.includes('gemini-2.0-flash')) {
} else if (model.model.includes('gemini-3.1-flash-lite')) {
expect(response.usageMetadata!.promptTokenCount).to.not.equal(0);
expect(response.usageMetadata!.candidatesTokenCount).to.not.equal(0);
expect(response.usageMetadata!.totalTokenCount).to.not.equal(0);
Expand Down
2 changes: 1 addition & 1 deletion packages/ai/src/constants.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,4 +39,4 @@ export const DEFAULT_FETCH_TIMEOUT_MS = 180 * 1000;
/**
* Defines the name of the default in-cloud model to use for hybrid inference.
*/
export const DEFAULT_HYBRID_IN_CLOUD_MODEL = 'gemini-2.5-flash-lite';
export const DEFAULT_HYBRID_IN_CLOUD_MODEL = 'gemini-3.1-flash-lite';
Loading