Function: google()

google(modelId, config?): ModelConfig

Defined in: packages/core/src/providers/factories.ts:517

Configure Google (Gemini) provider explicitly.

Google's Gemini family is served through the Generative Language API at https://generativelanguage.googleapis.com/v1beta and routed via @ai-sdk/google. Model IDs follow the gemini-* prefix (gemini-2.5-pro, gemini-2.5-flash, gemini-2.0-flash, gemini-1.5-pro, gemini-1.5-flash, …). Auto-detection routes any gemini-* model here; the factory is for explicit config (custom apiKey, baseUrl, reasoning, headers, fetch).

Multimedia: Gemini 1.5+/2.x is multimodal-first — images, video frames, audio, and PDF/file inputs are accepted natively through the AI SDK content-part path.

Reasoning ("thinking"): Gemini 2.5 models expose thinking budgets via thinkingConfig. reasoning.effort maps to thinkingBudget (low=2048, medium=8192, high=24576), or pass reasoning.budgetTokens directly for fine-grained control.

API key env vars (checked in order): GEMINI_API_KEY, GOOGLE_GEMINI_API_KEY, GOOGLE_GENERATIVE_AI_API_KEY.

Parameters

modelId

string

Gemini model ID, e.g. 'gemini-2.5-flash'

config?

ProviderConfig & object

Optional configuration (provider settings + reasoning).

Returns

ModelConfig

Examples

const bot = new Baleybot({ model: google('gemini-2.5-flash') });

const bot = new Baleybot({
  model: google('gemini-2.5-pro', { reasoning: { effort: 'high' } })
});

const bot = new Baleybot({
  model: google('gemini-2.5-pro', { reasoning: { budgetTokens: 16384 } })
});

const bot = new Baleybot({
  model: google('gemini-2.5-flash', { apiKey: 'AI...' })
});