Skip to content

feat: use google/genai sdk for streaming gemini & vertex responses #8907

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
wants to merge 6 commits into
base: main
Choose a base branch
from
+240 −872
Open

feat: use google/genai sdk for streaming gemini & vertex responses #8907

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
b0ee207
add @google/genai
uinstinct Nov 27, 2025
387374e
refactor Gemini adapter to use the sdk
uinstinct Nov 27, 2025
f8ad80e
refactor vertexai adapter to use genai sdk
uinstinct Nov 27, 2025
c96c5c7
use openai adapter for gemini and vertex in core
uinstinct Nov 27, 2025
42d56e7
Merge branch 'main' into gemini-sdk
uinstinct Nov 27, 2025
77d63db
reinit package-lock
uinstinct Nov 27, 2025
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
10 changes: 10 additions & 0 deletions core/llm/llms/Gemini.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,6 +13,7 @@ import { safeParseToolCallArgs } from "../../tools/parseArgs.js";
import { renderChatMessage, stripImages } from "../../util/messageContent.js";
import { extractBase64FromDataUrl } from "../../util/url.js";
import { BaseLLM } from "../index.js";
import { LlmApiRequestType } from "../openaiTypeConverters.js";
import {
GeminiChatContent,
GeminiChatContentPart,
Expand Down Expand Up @@ -41,6 +42,15 @@ class Gemini extends BaseLLM {
maxEmbeddingBatchSize: 100,
};

protected useOpenAIAdapterFor: (LlmApiRequestType | "*")[] = [
"chat",
"embed",
"list",
"rerank",
"streamChat",
"streamFim",
];

// Function to convert completion options to Gemini format
public convertArgs(options: CompletionOptions): GeminiGenerationConfig {
// should be public for use within VertexAI
Expand Down
10 changes: 10 additions & 0 deletions core/llm/llms/VertexAI.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,6 +5,7 @@ import { ChatMessage, CompletionOptions, LLMOptions } from "../../index.js";
import { renderChatMessage, stripImages } from "../../util/messageContent.js";
import { BaseLLM } from "../index.js";

import { LlmApiRequestType } from "../openaiTypeConverters.js";
import Anthropic from "./Anthropic.js";
import Gemini from "./Gemini.js";

Expand All @@ -23,6 +24,15 @@ class VertexAI extends BaseLLM {

private clientPromise: Promise<AuthClient | void>;

protected useOpenAIAdapterFor: (LlmApiRequestType | "*")[] = [
"chat",
"embed",
"list",
"rerank",
"streamChat",
"streamFim",
];

/*
Vertex Supports 3 different URL formats
1. Standard VertexAI: e.g. https://{location}-aiplatform.googleapis.com/v1/projects/{project}/locations/{location}/publishers/google/models/{model}:streamGenerateContent
Expand Down
Loading
Loading