Supported Providers
Fermi supports cloud APIs and local inference servers. Use fermi init to configure any combination.
Provider Table
Section titled “Provider Table”| Provider | Models | Auth |
|---|---|---|
| Anthropic | Claude Fable 5.1, Opus/Sonnet/Haiku 5.5, active earlier models | ANTHROPIC_API_KEY |
| OpenAI | GPT-6.1 Sol, GPT-6 Astra/Sol/Luna, GPT-5.6 Sol/Terra/Luna, active earlier models | OPENAI_API_KEY or OAuth |
| GitHub Copilot | Current Claude/GPT, Gemini Flash, Kimi K3, MAI-Code; filtered by account catalog | /copilot device-flow login |
| DeepSeek | V4.1 Flash (vision), V4 Pro (0813, text) | Managed slot (FERMI_DEEPSEEK_*) |
| Kimi / Moonshot | K3, K2.7 Code/Highspeed, K2.6; separate Coding IDs | Managed slots (FERMI_KIMI_*) |
| MiniMax | M3.1 Flash Preview (M Plan), M3, M2.7/2.5 + Highspeed | Managed slots (FERMI_MINIMAX_*) |
| GLM / Zhipu | 5.3, 5.3 Flash, 5.2; FlashX on standard API only | Managed slots (FERMI_GLM_*) |
| Xiaomi (MiMo) | V2.6 Pro/Flash/UltraSpeed; V2.5 retires 2026-10-21 | Managed slot (FERMI_XIAOMI_*) |
| Qwen / DashScope | 3.8 Max/Flash/Omni, 3.7 Plus/Flash, regional open models | Managed slots (FERMI_QWEN_*) |
| OpenRouter | Multi-vendor curated presets (Claude, GPT, Kimi, MiniMax, GLM, DeepSeek, Qwen, Xiaomi) + any custom model | OPENROUTER_API_KEY |
| Ollama | Any local model (dynamic discovery) | — |
| oMLX | Any local MLX model (dynamic discovery) | — |
| LM Studio | Any local GGUF model (dynamic discovery) | — |
Cloud vs. Local
Section titled “Cloud vs. Local”Cloud providers require either an API key or an OAuth login. The init wizard prompts for keys and stores them in ~/.fermi/.env. Kimi, MiniMax, GLM, DeepSeek, Xiaomi, and Qwen use Fermi-managed internal slots. GitHub Copilot uses its own device-flow OAuth via /copilot. OpenAI (ChatGPT Login) stores OAuth tokens in ~/.fermi/state/oauth.json.
Local providers (Ollama, oMLX, LM Studio) connect to a server on your machine. No API key needed. During fermi init, the wizard queries the server’s model endpoint to discover available models.
Switching at Runtime
Section titled “Switching at Runtime”Use /model during a session to switch between any configured model. For providers with missing keys, selecting a model can prompt you to import or paste the key on the spot.
Use /tier to assign models to high/medium/low tiers for sub-agents.
See Model Switching for details.
Known Limitations
Section titled “Known Limitations”Third-party coding plans (Kimi-Code, GLM-Code) use whitelist-based access control. Unless your account has explicit access, these endpoints will reject requests. Standard API endpoints work normally.
Setup Guides
Section titled “Setup Guides”- Cloud Providers — Anthropic, OpenAI, DeepSeek, Kimi, GLM, MiniMax, Xiaomi, Qwen, OpenRouter
- GitHub Copilot — Use your GitHub Copilot subscription
- Local Providers — Ollama, oMLX, LM Studio
- ChatGPT OAuth Login — Use your ChatGPT account instead of an API key