Dolphin LLM: switch your client in three lines

Base URL and authentication

Our API is designed as a drop-in replacement for standard OpenAI integrations. You do not need a new SDK or a complex configuration file. Simply update the base_url in your client to point to our endpoint and provide your API key in the standard Authorization header.

Sign up for a free account to receive your key immediately. There is no card required for the trial credit, making it easy to verify connectivity. The base URL for all requests is https://new.dolphinllm.com/v1. This ensures that any tool, library, or application already configured for OpenAI can route traffic to our model with minimal friction.

Base URL

https://new.dolphinllm.com/v1

Auth header

Authorization: Bearer YOUR_API_KEY

Model ID

uncensored

First request

Test your connection with a simple chat completion request. This confirms that your key is valid and the endpoint is responsive. The request body follows the standard OpenAI chat completions schema. Replace YOUR_API_KEY with your actual key.

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

curl https://new.dolphinllm.com/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "uncensored",
    "messages": [{"role": "user", "content": "Write a blunt product review of a cheap VPN."}]
  }'

This endpoint supports streaming and tool calling, but this basic request demonstrates the core text-in, text-out capability. If you receive a response, your client is successfully connected to the Dolphin LLM service.

Python SDK

Use the official openai Python library to interact with the API. Configure the client with the custom base URL and your API key. This approach works with any code already written for OpenAI, allowing you to switch providers without rewriting your logic.

from openai import OpenAI

client = OpenAI(base_url="https://new.dolphinllm.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)

The model ID is uncensored. Pass the full conversation history in messages.

Node SDK

For Node.js applications, the openai npm package handles the integration. Set the baseURL to our endpoint and provide your key. This ensures compatibility with existing Node.js projects that rely on OpenAI's structure.

import OpenAI from "openai";

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

The API supports standard parameters like max_tokens and temperature. Adjust these to control the creativity and length of the output.

Streaming responses (SSE)

Enable streaming by setting the stream parameter to true in your request body. The API returns Server-Sent Events (SSE), allowing you to process tokens as they are generated. This reduces perceived latency for end users.

stream = 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)

The stream uses standard OpenAI event types, so most SSE parsers work with it. Token usage arrives in the last data frame before data: [DONE].

Rate limits and constraints

The API enforces a limit of 300 requests per minute and 8 parallel requests per key (trial keys: 60 and 2). The request body size is capped at 8 MB. If you exceed these limits, you will receive a 429 error.

Authentication errors return 401 if the key is missing or invalid. Payment errors return 402 if your balance is empty. The context window is 64,000 tokens for both prompt and completion combined; keep your inputs within this limit.