Call Google's Gemini family through the same OpenAI-compatible /v1 endpoint as GPT and Claude. Useful when you want to A/B models without wiring up a separate Google client.
| Feature | Vynex API | direct Google |
|---|---|---|
| Gemini 3.x Pro / Flash | ✓ | ✓ |
| OpenAI-compatible interface | ✓ | Google SDK |
| Same key for GPT & Claude | ✓ | ✗ |
| Streaming + tool calling | ✓ | ✓ |
| Per-token pricing | ✓ | ✓ |
Yes. Set base_url to Vynex and pass a Gemini model name. The interface is OpenAI-compatible, so you can mix Gemini with GPT and Claude without multiple SDKs.
The Gemini 3.x and 2.5 family including Pro and Flash variants. See /pricing for the current list.