Features & Capabilities

What AI models are supported?

openFluxHub 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: VEO3, Kling, etc.
• Audio Models: Whisper-1, TTS, etc.

Please check the API Manual (/docs) for specific supported models.

How compatible is it with OpenAI API?

openFluxHub 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.openfluxhub.com` 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.