Features

This page answers common questions about model support, OpenAI compatibility, and response differences.

Which AI models are supported?

The platform provides one API surface for multiple model and data APIs. Common capabilities include:

• Text and chat models
• Image generation models
• Video generation models
• Twitter and other data APIs

Available models are determined by the console or /v1/models. Different API Keys may see different models because of model permissions or configuration.

How compatible is it with the OpenAI API?

If your client supports custom Base URL and API Key, you can usually integrate with small changes.

Common OpenAI-compatible entries include:

• POST https://api.requestbeam.com/openai/v1/chat/completions
• POST https://api.requestbeam.com/openai/responses

Notes:

• Use model names from the platform model list; do not assume every official model name is available.
• Common OpenAI-style fields are supported where available, but exact support depends on the selected model and available features.
• Gemini, Claude, and other model-specific features may not fully map to OpenAI format.

Why do API results differ from official web apps?

This is normal and does not necessarily indicate a failed call. Official web apps often add extra engineering layers such as system prompts, history handling, safety policies, and UI-side post-processing.

API calls are closer to raw model behavior. Applications usually need to manage:

• System prompts
• Conversation history
• Output format constraints

For fair comparisons, align model ID, system prompt, context, and parameters, then compare multiple runs instead of one output.