gptossToapi

Reverse gpt-oss.com to an openai compatible API

GPT-OSS API Proxy

Overview

This project is an API proxy that provides a compatible interface with OpenAI's API specification for interacting with GPT-OSS models. It supports both streaming and non-streaming responses for chat completions.

Features

• OpenAI API compatible endpoints
• Support for multiple GPT-OSS models
• Streaming response support (Server-Sent Events)
• CORS enabled
• Simple deployment with Deno

Installation

1. Ensure you have Deno installed (https://deno.land/)
2. Clone this repository
3. Run the server:

deno run --allow-net main.ts

API Endpoints

GET /

• Returns basic server status and available endpoints

GET /v1/models

• Lists available models
• Response format:

{
  "object": "list",
  "data": [
    {
      "id": "gpt-oss-120b",
      "object": "model",
      "created": 1234567890,
      "owned_by": "gpt-oss"
    },
    {
      "id": "gpt-oss-20b",
      "object": "model",
      "created": 1234567890,
      "owned_by": "gpt-oss"
    }
  ]
}

POST /v1/chat/completions

• Processes chat completion requests
• Supports both streaming and non-streaming responses
• Required parameters:
  • messages: Array of message objects
  • model: Model ID (default: "gpt-oss-120b")
  • stream: Boolean to enable streaming

Example Request:

{
  "model": "gpt-oss-120b",
  "messages": [
    {"role": "system", "content": "You are a helpful assistant."},
    {"role": "user", "content": "Hello!"}
  ],
  "stream": false
}

Configuration

Available Models

The proxy currently supports these models:

• gpt-oss-120b
• gpt-oss-20b

The default model is gpt-oss-120b.

Environment Variables

No environment variables are currently required, but you may want to add:

• PORT - To specify the server port
• CORS_ORIGIN - To restrict CORS access

Error Handling

The API returns appropriate HTTP status codes and error messages in JSON format for:

• Invalid requests (400)
• Method not allowed (405)
• Not found (404)
• Server errors (500)

Deployment

This server is designed to run with Deno. For production deployment, consider:

• Using Deno Deploy
• Running behind a reverse proxy (Nginx, Caddy)
• Implementing rate limiting
• Adding authentication