
Windows Agent Arena is Microsoft’s research environment for testing computer-use agents on realistic Windows tasks.
Researchers can run agents against more than 150 Windows tasks covering browser work, office applications and system settings. The project includes Navi as a reference agent and supports local evaluation or parallel runs through Azure Machine Learning.
Setup requires Docker, Python, model credentials and a prepared Windows 11 virtual machine. Researchers choose the agent, tasks and difficulty, then inspect the resulting actions and task outcomes. It is a benchmark environment rather than a ready-made personal desktop assistant.
The linked September 2024 video introduces the original research release; current setup instructions are in the repository.
Product overview
Windows Agent Arena is Microsoft’s research environment for testing computer-use agents on realistic Windows tasks.
Information as of October 19, 2026
31/100
Our subjective rating of how independently the product carries out its documented tasks, considering the human setup, input, approval and review it needs. This is not a measured success rate.
Loading Community Opinions...
Show prospects that Windows Agent Arena has a public place where they can check product details, pricing, ratings, and reviews.
Build confidence
Give buyers a third-party profile to explore.
Reduce hesitation
Put validation beside your strongest CTA.
Earn discovery
Every badge links prospects to your listing.
Choose your style
Preview it, then copy the complete embed code.
Shows buyers where to validate your product, pricing, and reputation.
Plain HTML. No signup, script, or maintenance required.
For ai agent research labs
Exploring Windows Agent Arena? For AI agent research labs, your own AI agent can map research evidence, run reproducible experiments, and prepare research artefacts. Start with a setup tailored to this work.
Tailored starting setup. Chat on our site. See the workflow and monthly plan.