What is Janitor AI?
Janitor AI is a character creation and immersive chat platform that can use a custom proxy for compatible model requests.
This Janitor AI Proxy setup follows a publicly documented third-party integration pattern. Confirm Proxy Configurations are available in your account, make a short test request, and monitor tokens before starting a long conversation. The platform and the model API are separate services, so availability and billing must be evaluated independently. This guide relies on a public third-party Custom Proxy workflow because a stable official integration reference is unavailable; confirm that Proxy Configurations exist in the current account UI and monitor context growth during long chats.
This page is not a Janitor AI compatibility statement. We did not find an official Janitor AI guide for TokenHub. The steps below map the public JanitorAI Proxy tutorial published by NanoGPT, changing only the provider URL, key, and model ID to TokenHub. If the interface changes, follow the current Janitor AI product.
Janitor AI differs from most apps here: its Proxy URL field takes the full request URL https://us-api.tokenhub.com/v1/chat/completions, not only /v1.
Follow the seven steps in the third-party tutorial
- Open API Settings in Janitor AI.
- Choose Proxy as the provider.
- Under Proxy Configurations, click + Add.
- Enter
TokenHubfor Configuration Name. - Enter an exact TokenHub model ID for Model Name.
- Enter
https://us-api.tokenhub.com/v1/chat/completionsfor Proxy URL and your TokenHub API key for API Key. - Click Apply, then confirm that this TokenHub Proxy is marked Active.
Unlike Open WebUI or SillyTavern, this third-party workflow uses the complete Chat Completions path. If your Janitor AI account has no Proxy or Proxy Configurations entry, stop; do not force these values into a different provider form.
Send a real message to verify the Active configuration
Send a short message and inspect TokenHub logs for /v1/chat/completions, the model ID, and token usage. A request in those logs is the reliable confirmation that the Active configuration is TokenHub.
Troubleshoot in the third-party tutorial's order
- No request leaves the page: confirm you clicked Apply and that the TokenHub Proxy is Active.
PROXY ERRORor 404: copy the current model ID again and confirm that Proxy URL is the complete Chat Completions URL.- 401: paste the TokenHub API key again and remove surrounding whitespace.
- Failure after changing models: Janitor AI does not update an old model ID automatically; edit the configuration and Apply again.
- Current UI differs from the tutorial: treat the integration as unverified and follow the current Janitor AI interface; do not infer official compatibility.
Third-party reference
Choose a compatible Janitor AI model
Janitor AI Custom Proxy FAQ
Does Janitor AI officially document this TokenHub setup?
No official Janitor AI setup source was identified during this audit. This page is based on a publicly documented third-party Proxy integration and should be verified in the current UI before use.
Should I enter the TokenHub API root or the complete Proxy URL?
Use the complete Chat Completions request URL ending in /v1/chat/completions. The Proxy field documented for Janitor AI does not take the API root alone.
What should I check when the Proxy does not return a reply?
Confirm the Proxy configuration is Active, the model ID is exact, the API key has no extra spaces, and the TokenHub request log receives a Chat Completions request.
Is Janitor AI usage through TokenHub free?
TokenHub model requests are billed using the selected model's current input, output, and applicable cache rates. Janitor AI platform pricing is a separate service layer.
How do I fix a Janitor AI Proxy 404?
Use the complete /v1/chat/completions endpoint, verify the key and model ID, make the Proxy active, and retry with a short message.
References
This setup is based on TokenHub and official documentation, last verified on 2026-09-08.