Chat Completions API Quickstart

Conversational AI models including Claude Sonnet 4. Chat models use the OpenAI-compatible gateway — not the async Jobs API.

Capabilities: chat completions
Playground: /playground/claude-chat
Estimated cost: from 5 credits per request (token metering may apply — see pricing)

Authentication

Authorization: Bearer YOUR_API_KEY
Content-Type: application/json

Quick Start

Kie market chat models (codex, gpt-5-*, gemini-*, catalog claude-*) should send "provider": "kie". If omitted, the gateway still routes known Kie catalog ids to Kie (not OpenRouter).

curl -X POST "https://modkie.com/api/gateway/chat/completions" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "kie",
    "model": "codex",
    "messages": [
      { "role": "user", "content": "Summarize the benefits of a unified AI API." }
    ]
  }'

OpenRouter-style org/model ids (e.g. anthropic/claude-sonnet-4) still default to OpenRouter when provider is omitted.

Response follows the OpenAI chat completions format with choices[].message.content.

Streaming

Add "stream": true to receive Server-Sent Events:

{
  "model": "anthropic/claude-sonnet-4",
  "stream": true,
  "messages": [{ "role": "user", "content": "Hello" }]
}

Key Parameters

Billing

Chat requests are billed synchronously based on model credit rates and token usage. Check balance with Get Remaining Credits.

Next Steps

• Market Models — all available LLMs
• Get Remaining Credits
• Getting Started — platform overview

Support

• Discord · Telegram · [email protected]