Guides
OpenAI compatibility
Use the OpenAI SDK's Chat Completions interface with a custom base URL. Support is explicit for each endpoint and model capability.
Supported endpoints
| Endpoint | Support |
|---|---|
| GET /v1/models | Enabled catalog with verified metadata. |
| GET /v1/models/{model_id} | One approved model. |
| POST /v1/chat/completions | Text completions, verified tool calling, and SSE. |
Responses, embeddings, images, audio, files, fine-tuning, and Anthropic Messages endpoints are not implemented. Clients that require those protocols need a different supported mode.
Request fields
| Field | Accepted values |
|---|---|
| model, messages | Required model ID and 1–128 messages. |
| roles | system, developer, user, assistant, tool. |
| max_tokens or max_completion_tokens | One output limit; also bounded by the model. |
| temperature, top_p | 0–2 and 0–1 respectively. |
| stream, stream_options | Boolean streaming and optional include_usage. |
| stop | A string or up to four strings. |
| tools, tool_choice | Verified function tools; up to 32 definitions. |
| n | Exactly 1. |
Unknown parameters and unsupported message fields are rejected rather than silently ignored. Message content is text; image or audio content arrays are not accepted.
Client setup
POST /v1/chat/completions
curl "https://uncensoredlabs.dev/v1/chat/completions" \
-H "Authorization: Bearer $UNCENSOREDLABS_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"YOUR_APPROVED_MODEL_ID","messages":[{"role":"user","content":"Hello"}],"max_tokens":256}'