Since the servers are always “exhausted,” could it be possible to integrate local Ollama models for both offline and a more stable agent experience, or to just use your own OpenAI, Gemini, Anthropic, etc., API key?
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| Servers busy error | 0 | 155 | April 9, 2026 | |
| Antigravity add your own API Keys/Models | 1 | 2028 | March 30, 2026 | |
| [feature request] Model quota relaxation: Allow AntiGravity to connect local models or other models via OAuth | 1 | 254 | March 2, 2026 | |
| ChatGPT as a selectable model | 2 | 881 | January 27, 2026 | |
| Antigravity Models Are Not Working - Agent terminated due to error | 7 | 761 | April 20, 2026 |