LibreChat
Self-hosted multi-model chat UI.
LibreChat exposes any OpenAI-compatible host through `endpoints.custom`. Add a FreeTheAi block, set the env var, and the model picker auto-fills from /v1/models.
Chat completions Streaming Tool calling
Connect LibreChat to FreeTheAi.
- 1
Edit librechat.yaml
In your LibreChat install, add a custom endpoint block under `endpoints.custom`.
yaml endpoints: custom: - name: 'FreeTheAi' apiKey: '${FREETHEAI_API_KEY}' baseURL: 'https://api.freetheai.xyz/v1' models: default: ['bbg/zai-org/GLM-5.1', 'rev/claude-sonnet-4.5'] fetch: true titleConvo: true titleModel: 'bbg/deepseek-ai/DeepSeek-V3.2' modelDisplayLabel: 'FreeTheAi' - 2
Add the env var
Set FREETHEAI_API_KEY in your LibreChat .env file.
bash FREETHEAI_API_KEY=fta_PASTE_YOUR_FREETHEAI_KEY - 3
Restart LibreChat
Restart the container or process. The FreeTheAi endpoint will appear in the chat dropdown.
Models that pair well with LibreChat.
-
bbg/zai-org/GLM-5.1 -
rev/claude-sonnet-4.5 -
bbg/deepseek-ai/DeepSeek-V3.2
See the full live catalog at /models. Aliases are stable; pricing on the free tier stays $0.
Common gotchas.
- Use `${FREETHEAI_API_KEY}` (single brace) for env-var values, not the secrets syntax used in some other configs.
- If a model rejects extra OpenAI params, add them to `dropParams` (e.g. `dropParams: ['frequency_penalty','presence_penalty']`).
- `fetch: true` auto-populates from FreeTheAi's catalog. Combine with `default` to surface preferred aliases first.