ferrin-google

Ferrin provider for Google Generative AI (Gemini): the generateContent language model (generate and stream, thinking, structured output, grounding sources, code execution, provider-executed tools), embeddings, Gemini image generation, text-to-speech, transcription through the Interactions API, Veo video operations, the Files API, batch generation and Live API sessions.

Version 0.2.0 adds GoogleProvider::interactions (generation, incremental streaming, background resume and resource start/get/cancel), plus streaming Live transcription and speech translation behind the realtime feature. Local fixture and WebSocket tests verify translation; real Google responses remain PV-031.

Part of the Ferrin workspace. Design: docs/01-architecture/17-provider-implementation-guide.md; capability matrix, provider options and metadata: docs/providers/google.md.

use ferrin_google::GoogleSettings;
use ferrin_google::create_google;

# fn main() -> Result<(), ferrin_spec::error::ProviderError> {
// Reads GOOGLE_GENERATIVE_AI_API_KEY on the first request.
let provider = create_google(GoogleSettings::default())?;
let model = provider.language_model("gemini-2.5-flash");
# let _ = model;
# Ok(())
# }

Testing

Tests live in tests/suite/ and replay hand-authored fixtures from tests/fixtures/ through ferrin_testing::FixtureServer; no network access or API key is needed:

cargo nextest run -p ferrin-google

License

Apache-2.0. See LICENSE and NOTICE. Portions of this crate are derived from the Vercel AI SDK (Apache-2.0); the crate and module documentation carry the attribution.