Quickstart

Before You Begin

You'll need an LLM247 account and a funded balance. The whole process takes less than five minutes.

1. Register Your Account

Visit llm247.net and create a free account. Email and password is all that's required — no payment details needed to sign up.

LLM247 runs on a prepaid credit system. Navigate to Billing inside the console and deposit an initial amount. Ten dollars is more than enough to explore every available model. Your credit never expires.

3. Create an API Key

Open API Keys in the console sidebar and click New Key. Give it a name, then copy the generated value immediately — it will only be shown once. Keys follow the format sk-llmaai-....

Keep your key private. Treat it the same way you would a database password — never check it into version control.

4. Send a Request

Paste your key into the command below and run it:

curl https://api.llm247.net/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kimi-k2.6",
    "messages": [
      {
        "role": "user",
        "content": "In one sentence, what is LLM247?"
      }
    ]
  }'

5. Read the Response

A successful call returns a JSON payload in the standard chat completions format:

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "model": "kimi-k2.6",
  "choices": [
    {
      "index": 0,
      "message": {
        "role": "assistant",
        "content": "LLM247 is a unified API gateway that lets you access multiple AI models through a single endpoint."
      },
      "finish_reason": "stop"
    }
  ],
  "usage": {
    "prompt_tokens": 16,
    "completion_tokens": 20,
    "total_tokens": 36
  }
}

The usage object tells you exactly how many tokens were consumed — and therefore how much was deducted from your balance.

Try Other Models

The same endpoint serves every supported model — just change the model field. A few popular slugs to experiment with:

gpt-5.4, gpt-5.5 — OpenAI's latest GPT-5 family
claude-opus-5, claude-opus-4.8 — Anthropic's frontier Opus reasoning models
gemini-3.5-flash, gemini-3.1-flash-lite — Google's Gemini 3 line
deepseek-v4-pro, kimi-k2.6, minimax-m3 — high-context, low-cost open-weight options
veo-3.1 — video generation (priced per generation, not per token)

See the Models reference for the full catalog with pricing.

Where to Go Next

Models — browse all 47 supported models with slugs, context windows, and pricing
Billing — understand exactly how balance deductions are calculated
OpenAI SDK — use LLM247 from Python or Node.js in minutes
Cursor IDE — wire LLM247 into your code editor