API
Your Base URL instead of api.openai.com
Same OpenAI SDK, different host: Base URL https://uncensored.monster/v1 and a key from the API button in @uncensobot. Protocol is Chat Completions only. Anthropic Messages, OpenAI Responses, embeddings, and FIM are not served.
Do not append /chat/completions to the Base URL — the client adds that suffix. Key path: how to start. Spec: OpenAI Chat Completions.
Why swap the Base URL
Your client already knows chat.completions.create. You change three things: URL, key, model id. Default is z-ai/glm-5.3-flash-uncensored. An empty model still routes to GLM. The model picker in Telegram does not change the API.
There is no web cabinet. Key and balance live in the bot. GET /v1/models needs Authorization: Bearer. Chat is POST /v1/chat/completions.
What will fail
An Anthropic or Responses client returns 404 — wrong protocol, not a missing model. Status codes: 401, 402, 404. Cursor-only steps: custom Base URL in Cursor.
Client context window 1048576, output budget 16384 or your max_tokens. Full parameter list is in the docs; this page does not copy it.
FAQ
Can I keep api.openai.com and only swap the key?
No. A key from our bot is accepted only at https://uncensored.monster/v1.
Do I need an Anthropic Base URL?
No. Leave it empty. Only Chat Completions is served.
Where is the key?
API button in @uncensobot. Shown once — copy immediately.