GH Repository · GetStream
Vision-Agents
Open Vision Agents by Stream. Build voice and vision agents quickly with any model or video provider. Uses Stream's edge network for ultra-low latency.
- stars
- 8,129
- 30-day movement
- +10/day
- Related entries
- 62
- Connections
- 4
python/uvpythonaivision-aiagentsai-agentsvoice-aivideo-agentsagent-appagentic-airealtimesttttsvideo-aiPython
Vision-Agents is Stream's open-source Python framework for building voice and vision agents. It is model- and provider-agnostic, routing realtime audio and video through Stream's edge network for low latency.
It gives you a single framework for realtime voice/video agents without locking you into one model or video provider.
Use it to
- Build realtime voice agents with STT and TTS
- Create agents that see and reason over live video
- Swap in any LLM, STT, TTS, or video provider
- Prototype agentic video apps in Python with uv
- Set up Claude or Cursor rules for agent-assisted development
For Python developers building realtime voice and video AI agents
- Role
- agent-app
- Language
- Python
- Licence
- Apache-2.0
- Forks
- 683
- Open issues
- 2
- Last push
- 2026-09-16
- Latest release
- v0.1.0 · 2025-10-09
- Skills shipped
- 2
topicsvoice-aivision-aivideo-agentsrealtimeagentic-aistt-tts