Janitor AI
Janitor AI can send chats to any OpenAI-compatible chat completions endpoint through its proxy settings. Janitor needs the full endpoint URL, not just the base URL.
Placeholders:
$API_BASEishttps://api.abliteratedllm.dev/v1,$API_ORIGINishttps://api.abliteratedllm.dev, and$YOUR_KEYis your API key.
Add a configuration
- Open a chat on Janitor AI.
- Click the gear icon at the top right, choose API Settings, then Add Configuration.
- Fill in:
| Field | Value |
|---|---|
| Config Name | Any name, for example GLM abliterated |
| Model Name | glm-5.3-flash-abliterated, deepseek-v4.1-flash or deepseek-v4-pro, exactly |
| Proxy URL | $API_BASE/chat/completions |
| API Key | $YOUR_KEY, with no extra spaces |
| Custom Prompt | Optional |
- Click Add, then Save Settings, and refresh the page.
- Send a test message.
Good to know
- The URL must end in
/chat/completions. Janitor calls the URL as entered. - Max tokens. Trial accounts are capped at 4096 output tokens per message; paid accounts can use up to the model's
max_output_tokensfrom GET /v1/models (65536 for our abliterated models). - 18+ only. See our Acceptable use policy.
- Pick an abliterated model for adult roleplay. Models labelled Official DeepSeek API are DeepSeek's stock models and keep DeepSeek's own refusals. See Models.
Verified against Janitor AI docs on 2026-11-03: TL;DR Quickstart Proxy Instructions.