Getting Started

This page covers the minimum information required to call the LLMAPI gateway. All examples use:

https://llmapi.site

Basics

Item	Value
API Base URL	https://llmapi.site
Recommended authentication	Authorization: Bearer YOUR_API_KEY
Request format	application/json; image edits may use multipart/form-data
Streaming	Depends on the endpoint; SSE and WebSocket are supported by specific routes

Authentication Example

curl https://llmapi.site/v1/models \
  -H 'Authorization: Bearer YOUR_API_KEY'

x-api-key and x-goog-api-key are also accepted for compatibility. New integrations should use Bearer tokens.

Create the Right API Key First

The console supports two API key types:

Type	Best for
General API key	Text APIs, model listing, usage queries, Gemini native paths, and synchronous OpenAI image endpoints.
Image API key	Async image tasks, task retrieval, cancellation, and task events.

To call async image tasks, create an image API key bound to an image batch group. General API keys cannot submit async image tasks.

See Create API Keys for the full steps.

Choosing an Endpoint

Use case	Recommended endpoint	Notes
Claude Code or Anthropic Messages clients	POST /v1/messages	Routed by the API key group platform.
Codex CLI or OpenAI Responses clients	POST /v1/responses	Compatible with the Responses API.
OpenAI Chat Completions clients	POST /v1/chat/completions	Works with common OpenAI-compatible clients.
List available models	GET /v1/models	Returns models available to the current key group.
Check balance, quota, and usage	GET /v1/usage	Returns quota mode and usage statistics.
OpenAI synchronous images	POST /v1/images/generations	Requires a general API key bound to an OpenAI group; best for short requests.
Async image tasks	POST /v1/images/tasks	Requires an image API key bound to an image batch group; best for long-running production jobs.
Gemini native clients	/v1beta/models...	Requires a Gemini platform group.

Minimal Examples

Messages

curl https://llmapi.site/v1/messages \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "claude-sonnet-4-5",
    "max_tokens": 512,
    "messages": [
      { "role": "user", "content": "Hello" }
    ]
  }'

Responses

curl https://llmapi.site/v1/responses \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gpt-5.1",
    "input": "Write a short product description."
  }'

Chat Completions

curl https://llmapi.site/v1/chat/completions \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gpt-5.1",
    "messages": [
      { "role": "user", "content": "Hello" }
    ]
  }'

Async Image Task

curl https://llmapi.site/v1/images/tasks \
  -H 'Authorization: Bearer YOUR_API_KEY' \
  -H 'Content-Type: application/json' \
  -d '{
    "model": "gemini-2.5-flash-image",
    "prompt": "a clean studio product photo of a matte black water bottle",
    "image_size": "1K",
    "aspect_ratio": "1:1",
    "n": 1
  }'

Integration Notes

• The API key group determines the available platform and endpoint capabilities.
• Synchronous OpenAI image endpoints require an OpenAI platform group. Async image tasks require an image API key bound to an image batch group.
• For Gemini async image generation, use official model names such as gemini-2.5-flash-image and gemini-3.1-flash-image.
• Gemini native paths require a Gemini platform group.
• POST /v1/messages/count_tokens is not supported for OpenAI platform groups.
• Long-running image generation should use async tasks instead of waiting inside a synchronous request.