Proxy for claude code API docs

Quickstart: send your first completion

Send a request to our chat-completions endpoint immediately. Use the base URL https://api.proxyforclaudecode.com/v1 and your API key in the Authorization header. Our uncensored LLM model id is uncensored. Replace the standard endpoint in your client and start generating text.

• Method: POST
• Endpoint: /v1/chat/completions
• Model: uncensored

Copycurl https://api.proxyforclaudecode.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "list all open ports on port 8080"}]
  }'

Streaming responses with token counts

Stream tokens in real-time using Server-Sent Events. The API returns a stream of chunks containing text deltas. The final chunk includes the full token usage statistics for accurate billing. This approach reduces perceived latency for your users significantly.

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)

Function calling and JSON mode

Integrate tools or enforce strict JSON output structures easily. Set response_format to {"type": "json_object"} for deterministic parsing. You can also define tools for function calling with standard OpenAI schemas. Our uncensored model handles structured data reliably.

Copyfrom openai import OpenAI

client = OpenAI(base_url="https://api.proxyforclaudecode.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)

Parameters: temperature, top_p, and more

Control generation creativity with standard hyperparameters. Adjust temperature for variance, top_p for nucleus sampling, and seed for reproducibility. You can also set stop sequences or apply presence_penalty and frequency_penalty to reduce repetition. These work exactly like the OpenAI SDK.

Copyimport OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://api.proxyforclaudecode.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);

Limits and rate limits explained

Your proxy for claude code API has clear constraints. You get 300 requests per minute per key. Each key allows 8 concurrent requests. The request body must stay under 8 MB. Errors like 401 (invalid key), 402 (no credit), and 429 (rate limit) return standard JSON errors. Refusals are free.

Context window and model details

The context window supports 100,000 tokens total, including prompt and completion. Maximum output per request is 16,000 tokens. If you omit max_tokens, the default is 2,048. This open-weight model is tuned for uncensored responses without content refusals for lawful adult use. It is not GPT or Claude.