Gemini

fromGemini wraps the raw Gemini client — not a pre-configured model instance — since the adapter passes model through on every call rather than baking it in up front.

New SDK (@google/genai)

The current SDK's ai.models.generateContent({ model, ... }) takes model per call, which lines up directly with how fromGemini forwards it:

import { GoogleGenAI } from '@google/genai';
import { VernLLM, fromGemini } from 'vern-llm';

const ai = new GoogleGenAI({ apiKey: process.env.GEMINI_API_KEY });

const llm = new VernLLM({
  client: fromGemini(ai.models),
  model: 'gemini-2.5-flash',
});

Legacy SDK (@google/generative-ai)

The older SDK bakes the model into getGenerativeModel({ model }), so its generateContent doesn't accept a model field per call — fromGemini still adapts it since it only calls .generateContent, but the model you pass to VernLLM/per-call override is effectively ignored by the underlying client. Whatever model you picked at getGenerativeModel(...) time is what actually runs:

import { GoogleGenerativeAI } from '@google/generative-ai';
import { VernLLM, fromGemini } from 'vern-llm';

const genAI = new GoogleGenerativeAI(process.env.GEMINI_API_KEY);
const model = genAI.getGenerativeModel({ model: 'gemini-2.5-flash' });

const llm = new VernLLM({
  client: fromGemini(model),
  model: 'gemini-2.5-flash', // kept in sync manually — the legacy client ignores this per call
});