MiraBridge supports three AI providers, each with distinct strengths.
Model Comparison
| Model | Provider | Best For | Context Window | |-------|----------|----------|----------------| | Claude Sonnet 4.6 | Anthropic | Code generation, reasoning | 200K tokens | | GPT-5.4 | OpenAI | General tasks, speed | 128K tokens | | Gemini 2.5 Flash | Google | Fast responses, large context | 1M tokens |
Claude (Anthropic)
Claude excels at code generation, architectural reasoning, and following complex instructions. It is the default model for BUILD and ARCHITECT modes.
GPT (OpenAI)
GPT offers broad general knowledge and fast response times. It is well-suited for quick questions and FAST mode.
Gemini (Google)
Gemini Flash provides very fast responses with a massive context window. It handles large codebases well and is cost-effective for high-volume usage.
Automatic Failover
If one provider is unavailable, MiraBridge automatically falls back to an alternative provider. This ensures uninterrupted coding sessions.
Next Steps
- Bring Your Own Key โ use your own API keys
- AI Modes โ choose the right mode for your task