idosal/AgentLLM
Browser-native autonomous agent using open-source LLMs (WizardLM/Vicuna) via WebGPU for GPU-accelerated inference.
Not currently ranked — collecting fresh signals.
star history
AgentLLM runs autonomous agents entirely in the browser by embedding open-source LLMs. It leverages WebLLM and Chromium’s WebGPU API to perform GPU-accelerated inference client-side, eliminating server dependencies. The project is built upon AgentGPT’s architecture, uses LangChain for agent orchestration, and supports models like WizardLM and Vicuna for goal-oriented autonomous task completion.
Frequently asked
- What is idosal/AgentLLM?
- Browser-native autonomous agent using open-source LLMs (WizardLM/Vicuna) via WebGPU for GPU-accelerated inference.
- Is AgentLLM open source?
- Yes — idosal/AgentLLM is open source, released under the GPL-3.0 license.
- What language is AgentLLM written in?
- idosal/AgentLLM is primarily written in JavaScript.
- How popular is AgentLLM?
- idosal/AgentLLM has 468 stars on GitHub.
- Where can I find AgentLLM?
- idosal/AgentLLM is on GitHub at https://github.com/idosal/AgentLLM.