OpenRouter api Documentation

Quickstart: Connect with Python

Start by installing the official SDK and pointing it at our base URL.

Our API behaves like the standard OpenAI interface but routes to a single uncensored model. This approach eliminates proxy complexity while keeping your code portable.

• Install the library via pip.
• Set your API key in the environment.
• Define the base URL for our endpoint.

Copyfrom openai import OpenAI

client = OpenAI(base_url="https://api.apiopenrouter.com/v1", api_key="YOUR_KEY")

resp = client.chat.completions.create(
    model="uncensored",
    messages=[{"role": "user", "content": "Summarise this thread without softening it."}],
)
print(resp.choices[0].message.content)

Replace the placeholder key with your own credentials. You can send simple prompts or complex structured data.

Authentication and Keys

Generate a key on the dashboard and pass it in the Authorization header.

Each account supports one active key at a time. When you create a new key, the old one stops working immediately. This simplifies security management without requiring complex rotation logic.

Copycurl https://api.apiopenrouter.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "List the top 5 dependencies for a Go web server"}]
  }'

The request above uses the Bearer scheme. Ensure your key starts with sk- and contains no extra whitespace.

Chat Completions Endpoint

Send messages to the /v1/chat/completions endpoint using standard JSON.

The model accepts user and assistant roles. You can configure temperature, top-p, and stop sequences to control output style. Responses return token usage in the final chunk if streaming is enabled.

Copyimport OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.apiopenrouter.com/v1", apiKey: process.env.API_KEY });

const resp = await client.chat.completions.create({
  model: "uncensored",
  messages: [{ role: "user", content: "Draft a villain monologue for my game." }],
});
console.log(resp.choices[0].message.content);

Set max_tokens to control output length. If omitted, the default is 2,048 tokens. The model supports up to 16,000 tokens per request.

Streaming Responses

Enable streaming by setting stream: true in your request body.

Token usage details appear only in the last chunk of the stream. This allows real-time UI updates while preserving accurate billing data. Errors during streaming still report usage for charged tokens.

Copystream = client.chat.completions.create(
    model="uncensored",
    messages=[{"role": "user", "content": "Tell the story in second person."}],
    stream=True,
)
for chunk in stream:
    if chunk.choices and chunk.choices[0].delta.content:
        print(chunk.choices[0].delta.content, end="", flush=True)

Handle partial responses gracefully. The stream closes normally when the model finishes generation or hits the context limit.

Function Calling & JSON Mode

Use response_format: {type: "json_object"} for strict JSON outputs.

Define tools in the tools array and set tool_choice to enforce function calls. The model returns structured data suitable for programmatic processing. This works reliably with our uncensored weights.

Copycurl https://api.apiopenrouter.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "List the top 5 dependencies for a Go web server"}]
  }'

Combine function calling with streaming for interactive experiences. Validate the JSON output before passing it to your application logic.

Limits and Quotas

Monitor your usage against the per-key limits to avoid interruptions.

Each key allows 300 requests per minute and 8 concurrent requests. The request body size limit is 8 MB. If you exceed these limits, the API returns a 429 status code. Authentication errors return 401, and insufficient credit returns 402.

Copycurl https://api.apiopenrouter.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "List the top 5 dependencies for a Go web server"}]
  }'

The context window is 100,000 tokens total. Errors do not consume credit. Refusals for minor content violations also remain free.