Google Gemini Guide
Gemini support uses Google's OpenAI-compatible endpoint. This lets 3sparks Chat talk to Gemini with the same OpenAI-style message format used by many other providers.
Add the connection
- Open 3sparks Chat and go to Connections.
- Tap + and choose Google - OpenAI compatible API.
- Leave the name blank for a generated name, or enter your own.
- Confirm the base URL is
generativelanguage.googleapis.com. - Confirm the route prefix is
/v1beta/openai/. - Paste your Google API key into API Token. Tokens are stored in Keychain.
- Tap Create Connection.
- Tap Fetch Models, choose a default model, and tap Done.
Vision support
Gemini models that support image input can be used with camera captures, photos, and pasted images. If an image request fails, switch to a Gemini model with vision support and retry.
Troubleshooting
- Confirm the Google API key has access to Gemini API models.
- Keep
/v1beta/openai/as the route prefix for Google's compatible endpoint. - If no models appear, check API access in Google AI Studio and try Fetch Models again.
- Provider model names change over time, so use the fetched list rather than old examples.