Skip to main content

​
What AI models are supported?

APIMart supports various mainstream AI models, including:
  • Text Models: GPT-5, Claude, Gemini, etc.
  • Image Models: GPT-4o-image, Gemini-2.5-Flash-Image-preview, etc.
  • Video Models: Sora2, VEO3, etc.
  • Audio Models: Whisper-1, TTS, etc.
Please check the API Documentation for specific supported models.

​
How compatible is it with OpenAI API?

APIMart is fully compatible with OpenAI API format, allowing you to:
  • Directly replace the API endpoint
  • Use the same request format
  • No need to modify existing code
  • Support all major OpenAI API features
Simply replace the original OpenAI API endpoint https://api.openai.com with https://api.apimart.ai for seamless migration.

​
Why do API responses differ from official web versions (like ChatGPT, Claude)?

While using the same underlying models, there are differences between official web versions and API calls: Official Web Versions include extensive engineering optimizations and enhanced features:
  • Built-in internet search for real-time information
  • Conversation memory that automatically links context
  • Built-in calculator and code execution environment
  • Pre-configured system prompts
  • Optimized user interface
API Calls provide raw model capabilities:
  • Only core reasoning and generation abilities
  • Developers must manually manage conversation history
  • Requires manual configuration of system prompts and context
  • Requires self-integration of tool calling (Function Calling)
Development Tip: When using the API, actively configure system prompts (system messages), maintain conversation history, and integrate tool calling features as needed to achieve an experience closer to the web version.