API reference

The UncensoredAI API

An OpenAI-compatible HTTP API. If your code already speaks to OpenAI, point it at https://uncensoredai.dev/v1 and swap the key.

Quickstart

Every request targets POST /v1/chat/completions with a JSON body. Set stream: true to receive server-sent events instead of a single JSON document.

chat completion (non-streaming)

const response = await fetch("https://uncensoredai.dev/v1/chat/completions", {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    Authorization: `Bearer ${process.env.UNCENSOREDAI_KEY}`,
  },
  body: JSON.stringify({
    model: "gpt-6-sol",
    messages: [
      { role: "system", content: "You are a concise assistant." },
      { role: "user", content: "Summarise SSE in one sentence." },
    ],
    temperature: 0.7,
  }),
});

const data = await response.json();
console.log(data.choices[0].message.content);

Authentication

Your 16-digit access key is the credential. Send it as a Bearer token on every request, formatted as sk-ua-4821-9053-7712-6634. We store only a hash; rotating it from the dashboard invalidates the old value immediately.

request header

Authorization: Bearer sk-ua-4821-9053-7712-6634

Chat completions

POST /v1/chat/completions accepts the standard OpenAI fields. The model value must be one of the 35 catalog ids.

Field	Type	Notes
model	string	Required. A catalog model id.
messages	array	Required. OpenAI message objects.
stream	boolean	Optional. Enable SSE streaming.
temperature	number	Optional, 0–2. Forwarded upstream.
max_tokens	integer	Optional. Caps completion length.
tools	array	Optional. Function/tool definitions.

Streaming

With stream: true the response is a text/event-stream. Each data: line carries a delta frame; the stream ends with data: [DONE]. The final content frame may include a usage object.

streaming

curl https://uncensoredai.dev/v1/chat/completions \
  -H "Authorization: Bearer sk-ua-4821-9053-7712-6634" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek-v4-pro-uncensored",
    "messages": [{ "role": "user", "content": "Count to five." }],
    "stream": true
  }'

List models

GET /v1/models returns the catalog in OpenAI format. It requires the same Bearer authentication.

GET /v1/models

curl https://uncensoredai.dev/v1/models \
  -H "Authorization: Bearer sk-ua-4821-9053-7712-6634"

200 OK

{
  "object": "list",
  "data": [
    { "id": "gpt-6-astra", "object": "model", "owned_by": "uncensoredai" },
    { "id": "glm-5-3-abliterated", "object": "model", "owned_by": "Z.ai" }
  ]
}

Usage & billing

Every billed request records prompt, cached and completion tokens. The cost is computed from the per-model rates on the pricing page and deducted from your balance. When the balance reaches zero, requests fail with a 402 until you top up.

Errors

Failures use the OpenAI error envelope: { "error": { message, type, code } }.

Code	Status	Meaning
invalid_api_key	401	The Bearer key is missing, malformed or revoked.
model_not_found	404	The requested model id is not in the catalog.
insufficient_quota	402	Balance has run out. Top up and retry.
upstream_unreachable	502	The upstream provider could not be reached.
upstream_not_configured	503	Upstream credentials are not set on the server.
invalid_request_error	400	The request body is malformed or missing fields.