Choose how to run this agent
The AI model (0.5 – 7.5 GB depending on selection) downloads to your browser on first run. Subsequent runs use the cached version. Your GPU does the work — nothing is sent to AgentOp's servers.
⚡
Local agents run on your GPU via WebGPU — you'll want Chrome or Edge on a machine with a graphics card, or an Apple Silicon Mac. This agent runs on-device only.
Loading agent…
All AI inference runs in your browser. AgentOp's server only serves the agent code. View agent details