LLM API GATEWAY
SWAGGER SPECS

API Specifications Overview

The LLM API Gateway OpenAPI specification defines a unified interface for interacting with multiple large language model providers. This specification abstracts provider-specific differences, allowing developers to integrate once and support multiple backends seamlessly.

Base URL

https://api.llm-gateway.io/v1

Authentication

All API requests require authentication using Bearer tokens. Include your API key in the Authorization header:

Authorization: Bearer YOUR_API_KEY

Content Type

All request and response bodies use JSON format:

Content-Type: application/json

Available Endpoints

The gateway exposes the following endpoints for LLM interactions:

Schema Definitions

The OpenAPI specification defines comprehensive schemas for all request and response objects:

{
  "model": "string (required)",
  "prompt": "string | array",
  "max_tokens": "integer",
  "temperature": "number (0-2)",
  "top_p": "number (0-1)",
  "n": "integer (default: 1)",
  "stream": "boolean",
  "stop": "string | array"
}

{
  "model": "string (required)",
  "messages": [{
    "role": "system|user|assistant",
    "content": "string"
  }],
  "temperature": "number",
  "max_tokens": "integer",
  "presence_penalty": "number",
  "frequency_penalty": "number"
}

Response Schema

{
  "id": "string",
  "object": "text_completion | chat.completion",
  "created": "integer (timestamp)",
  "model": "string",
  "provider": "string",
  "choices": [{
    "index": "integer",
    "text": "string (for completions)",
    "message": {
      "role": "string",
      "content": "string"
    } (for chat)",
    "finish_reason": "stop | length | content_filter"
  }],
  "usage": {
    "prompt_tokens": "integer",
    "completion_tokens": "integer",
    "total_tokens": "integer"
  }
}

Integration Examples

Import the OpenAPI specification into your preferred tools:

Swagger UI

# Download the spec
curl https://api.llm-gateway.io/openapi.yaml -o openapi.yaml

# Run Swagger UI locally
docker run -p 8080:8080 \
  -e SWAGGER_JSON=/spec/openapi.yaml \
  -v $(pwd)/openapi.yaml:/spec/openapi.yaml \
  swaggerapi/swagger-ui

Postman Import

# Import directly via URL
https://api.llm-gateway.io/openapi.json

# Or download and import the file
File -> Import -> Upload Files -> openapi.json

Generate Client SDK

# Using OpenAPI Generator
npx @openapitools/openapi-generator-cli generate \
  -i https://api.llm-gateway.io/openapi.yaml \
  -g python \
  -o ./client-sdk

# Available generators: python, java, typescript, go, ruby, etc.

Curl Example

curl -X POST https://api.llm-gateway.io/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4",
    "messages": [
      {"role": "user", "content": "Hello, how are you?"}
    ]
  }'

Provider Support Matrix

The gateway normalizes responses across different LLM providers: