Toolbox/Agent SDKs & frameworks/HarryHsing/OmniAgent

HarryHsing/OmniAgent

OmniAgent (ICML 2026): the first native omni-modal agent for active video perception — a 7B agent that beats Qwen2.5-VL-72B with 73% fewer frames on LVBench.

View on GitHub →
PythonApache-2.0agentaudio-visualicml2026llm-agentlong-video-understanding
Stars
80
Δ 7d
+3
Δ 30d
+6
Age
4 mo
Last push
32d ago
8375686053Jul 24Aug 23Oct 18
90-day star history

Where this sits in the catalog

Alternatives in Agent SDKs & frameworks

DeepSeek Harness: Everything is a Plugin.

⭐255.2k
+6,4947d
TypeScript
2 mo

The agent engineering platform.

⭐148.1k
+3777d
Python
4 yrs

Agent skills that give coding agents taste: art direction, reference images and a browser check instead of gen…

⭐99.6k
JavaScript
8 mo

Write HTML. Render video. Built for agents.

⭐68.0k
TypeScript
7 mo

Open-source personal AI assistant & Agent Harness. Plans tasks, runs tools and skills, self-evolves with memor…

⭐47.5k
+1007d
Python
4 yrs

Build, run, and manage agent platforms.

⭐42.8k
+1317d
Python
4 yrs