Features
Multimodal understanding (images and text)
Native function calling for tool use and API invocation
Code execution capabilities
Web search integration via function calling
OCR and document extraction
Visual question answering
Fine-tuning support via PEFT and full-parameter
Open-source weights (Apache 2.0)
Compatible with popular inference engines (vLLM, SGLang, Ollama)
Deployable on local hardware with consumer GPUs (e.g., RTX 4090)
Multi-agent orchestration with handoffs
Sandbox Agents for containerized long-running tasks
Agent-as-tool delegation (v0.15.0+)
Realtime Agents with gpt-realtime-2 voice support (v0.17.6+)
Human-in-the-loop mechanisms
Automatic session history management
Built-in tracing for debugging
Provider-agnostic LLM support (100+ models via LiteLLM)
Redis session support (optional, v0.17.6+)
Instructions, tools, and guardrails configuration
Jupyter notebook compatibility
Supports OpenAI Responses and Chat Completions APIs
