Janitor AI
Roleplay site that connects through a reverse proxy.
Janitor AI uses a proxy slot for any OpenAI-compatible endpoint. Drop FreeTheAi in as a proxy and pick the model you want.
Chat completions Streaming
Connect Janitor AI to FreeTheAi.
- 1
Open a chat
Pick a character on the home page and click `chat with` to land on the chat page.
- 2
Open API settings
Tap the three lines in the top right of the chat, then `API Settings`.
- 3
Switch to Proxy
In the API Settings panel, switch to the `Proxy` tab and click `+ New` to start a fresh config.
- 4
Fill the proxy form
Use these exact values. Janitor AI requires the full URL with /chat/completions.
text Name: freetheai Model: bbl/gpt-5.4-mini Proxy URL: https://api.freetheai.xyz/v1/chat/completions API Key: PASTE_YOUR_FREETHEAI_KEY - 5
Save
Click `Save`. Janitor AI will use FreeTheAi for new responses on this character.
Models that pair well with Janitor AI.
-
bbl/gpt-5.4-mini -
rev/claude-sonnet-4.5 -
glm/glm-5.1
See the full live catalog at /models. Aliases are stable; pricing on the free tier stays $0.
Common gotchas.
- Janitor AI sometimes shows a `Failed to fetch (unk)` error. That is upstream Janitor instability, not FreeTheAi - try Chub or SillyTavern instead, or wait it out.
- 401 means the key is wrong. Use `/resetkey` in the FreeTheAi Discord and paste only the key value, no quotes or markdown.
- Some models do not work for Janitor RP - try `rev/claude-sonnet-4.5` or `glm/glm-5.1` for the most reliable responses.