API Documentation

All endpoints require your OpenAI API key passed in the request body. No server-side authentication needed.

POST /api/chat

Send a message and receive a response from GPT models. Supports streaming.

Request Body

{
  "api_key": "sk-...",
  "messages": [
    {"role": "user", "content": "Hello!"}
  ],
  "model": "gpt-4o-mini",       // optional, default: gpt-4o-mini
  "temperature": 0.7,            // optional, 0–2, default: 0.7
  "stream": true                 // optional, default: false
}

Response (non-stream)

{
  "content": "Hello! How can I help you today?",
  "usage": {
    "prompt_tokens": 9,
    "completion_tokens": 12,
    "total_tokens": 21
  }
}

Response (stream)

Returns Server-Sent Events. Each event is a JSON object with a content field. Stream ends with data: [DONE].

POST /api/models

Fetch all GPT models available for your API key.

Request Body

{"api_key": "sk-..."}

Response

{"models": ["gpt-4o", "gpt-4o-mini", "gpt-3.5-turbo", ...]}

POST /api/image

Generate an image with DALL·E.

Request Body

{
  "api_key": "sk-...",
  "prompt": "A futuristic cityscape at sunset",
  "size": "1024x1024",     // 1024x1024 | 1792x1024 | 1024x1792
  "model": "dall-e-3"      // dall-e-3 | dall-e-2
}

Response

{
  "url": "https://oaidalleapiprodscus.blob.core.windows.net/...",
  "revised_prompt": "A neon-lit futuristic cityscape..."
}

POST /api/transcribe

Transcribe audio with Whisper. Send as multipart form data.

Form Fields

api_key: sk-...
audio: [audio file — mp3, mp4, wav, webm, m4a]

Response

{"text": "Transcribed text here..."}

Error Response Format

{"error": "Human-readable error message"}