Dolphin llm API Quickstart

Authentication & Base URL Setup

The API uses standard Bearer token authentication. You receive your key immediately after signing up with Google or an email address. Store this key securely in your environment variables. All requests must target the dedicated base URL for the uncensored model.

• Base URL: HTTPS://api.llmdolphin.com/v1
• Auth Header: Authorization: Bearer YOUR_API_KEY

Copycurl https://api.llmdolphin.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "Set variable x to 5"}]
  }'

The Chat Completions Endpoint

Send text prompts to the POST /v1/chat/completions endpoint. The model id is 'uncensored'. You can stream responses or receive them as full text blocks. The API accepts standard message formats with system, user, and assistant roles.

Copyfrom openai import OpenAI

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

Streaming Responses & Token Usage

Enable streaming by setting stream=true. The server returns Server-Sent Events (SSE). Token usage data appears in the final chunk of the stream. This allows you to track costs in real-time without waiting for the full response.

Copyimport OpenAI from "openai";

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

Function Calling & JSON Mode

The model supports function calling with tool definitions. You can also enforce JSON output using response_format. This ensures consistent parsing for your automation pipelines. Define tools in the messages array or request object.

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)

Parameters: Temperature, Top P, and Seed

Control creativity with temperature and diversity with top_p. Use seed for reproducible outputs. The model accepts standard parameters like stop sequences and presence penalty. Adjust these values to suit your specific use case.

Error Handling & Rate Limits

Handle 401 errors for invalid keys and 402 for insufficient credit. Rate limits are 300 requests per minute. The context window is 100,000 tokens. Errors and refusals do not consume credit. Check the response status for details.