ASUS has released the beta version of Zenni Claw, an agentic AI software designed to turn everyday prompts into guided, task-oriented workflows on supported ASUS devices, as the company pushes to make agentic AI more approachable for everyday users.
First introduced at Computex 2026, Zenni Claw is built around four principles: easy setup without complexity, ready-to-use assistant skills for work, travel and daily planning, flexible local and cloud AI processing options, and built-in safety guardrails. Unlike traditional AI chatbots that primarily respond to questions, the software is designed to help users move through entire tasks using guided workflows.
Supports Anthropic, OpenAI and Google models
The beta version supports various cloud AI models, including those from Anthropic, ChatGPT and Google Gemini, giving users flexibility to explore different AI-powered workflows depending on region and service availability. For local processing, Zenni Claw uses Gemma 4.
- Basic hardware requirements: Windows 11 64-bit or above, at least 16GB RAM, and 20GB of available storage.
- Running a local model requires at least 32GB RAM and a supported CPU platform such as Intel Panther Lake H12Xe or AMD Strix Halo.
- For a fuller local model experience, ASUS recommends an NVIDIA GeForce RTX 5090 GPU with 24GB or more VRAM.
Part of a broader AI PC push
The Zenni Claw beta is available as a free download for supported ASUS devices, with availability varying by region and device compatibility. ASUS said the release marks the next step in its broader effort to make AI PC experiences more practical, personal and accessible, and is encouraging users to share feedback during the beta period to help expand supported skills ahead of future versions.
