JanitorAI

JanitorAI connects to UberLLM through a proxy configuration in its API settings.

Before you start

An UberLLM account with credit, and an API key from Dashboard → Keys (ull_...).

Setup

Open a chat, click the gear icon, then API Settings.
Choose Proxy, then Add Configuration (shown as + Add under Proxy Configurations in some versions).
Fill in:
- Config Name: UberLLM
- Model Name: a model id, copied exactly, for example deepseek/deepseek-v4.1-flash
- Proxy URL: https://api.uberllm.dev/v1/chat/completions (the full address, with /chat/completions, no slash at the end)
- API Key: your ull_... key, with no spaces
- Custom Prompt: optional, can stay empty
Click Add, make sure the UberLLM configuration is the active one, click Save Settings, then refresh the page.

To change models later, edit Model Name in the same configuration, save and refresh.

Recommended models

Checked on 2026-10-30 through UberLLM.

Model name	Context	Notes
`deepseek/deepseek-v4.1-flash`	1,048,576	Fast and cheap
`z-ai/glm-5.2`	1,048,576	Good at long stories
`venice/uncensored-1.2`	128,000	Uncensored
`deepseek/deepseek-v4.1-flash-uncensored`	1,048,576	Uncensored; fewer providers serve it, so it can be busy

Uncensored text models are listed for every account. Turn on Text streaming in the chat's settings for long replies: a long reply without streaming can time out in JanitorAI before it arrives.

Common errors

What you see	Cause and fix
"A network error occurred, you may be rate limited or having connection issues"	Most often the URL: it must be exactly `https://api.uberllm.dev/v1/chat/completions`, no extra slash. Then save and refresh.
"Network Error" after Check Key/Model	JanitorAI's help says that button is unreliable. Save, refresh, and send a message instead.
`401`	Key missing, mistyped or with a space.
`404` `model_not_found`	Model Name mistyped. Copy the id from the catalog.
`402` `insufficient_credits`	Balance empty. Add credit under Dashboard → Credits.
Long replies fail after about 90 seconds	Turn on Text streaming.