Computer-use agents companies, products & suppliers
Agents that operate a desktop or OS GUI, not only a browser or API. Jobs: desktop gui automation with a model in the loop; os-level tool use; mixed app workflows on a vm.
What are Computer-use agents?
Agents that operate a desktop or OS GUI, not only a browser or API. Jobs: desktop gui automation with a model in the loop; os-level tool use; mixed app workflows on a vm.
What problems does it solve?
Legacy desktop software has no API and is still driven by people at a keyboard.
Typical business use cases
- Desktop GUI automation with a model in the loop
- OS-level tool use
- Mixed app workflows on a VM
Important capabilities
- Desktop/VNC runtime
- Screenshot plus actions
- VM isolation
What buyers should evaluate
- Isolation of the machine
- Credential vault
- Human approval for writes
Risks and governance considerations
An agent with a logged-in desktop image and standing privileges.
Procurement checklist
- VM tenancy
- Vault
- Recorded sessions
Relevant AI Trustmark assurance
AI Trustmark independent findings appear only when an assessment or certificate exists. Category membership does not imply verification.
Products and providers
Open a product profile for its catalogue details, or use the provider link for the company profile. Claimed listings appear first, then by name. Payment never changes the order.
Amazon Nova Act is an AWS service for building and managing fleets of AI agents that automate browser-based UI workflows (computer use) in production. Powered by a custom Nova 2 Li
By Anthropic
Anthropic is a US AI safety and research company that develops the Claude assistant family and related APIs. Claude Computer Use is Anthropic's model API, from Anthropic, PBC. Publ
Automation Anywhere is a US enterprise automation company selling robotic process automation and related AI agent software. Public marketing covers attended and unattended automati
By Browserbase
Stagehand is Browserbase's software product, from Browserbase, Inc. Published at browserbase.com/stagehand. Directory filing: Browser agents. Deployment options: SaaS and Self-host
By ByteDance
ByteDance is a technology company operating a range of content platforms that inform, educate, entertain and inspire people across languages, cultures and geographies. UI-TARS Desk
By Google
Gemini 3.6 Flash is a Flash-tier model in Google's Gemini API, built to give strong reasoning at higher speed and lower cost for real-world work. Google positions it for the agenti
By Induced
Public pages describe automating clicks and forms rather than wearable recovery. Induced is Induced's software product, from Buidl Inc. Published at induced.ai. Directory filing: C
By OpenAI
OpenAI is a US AI research company that sells hosted assistants, APIs and developer tools, including ChatGPT and the OpenAI API. ChatGPT Work is OpenAI's software product, from Ope
By Simular
Research-driven AI agents that operate real computers. Try Sai, your first robosecretary. Agent S2 is Simular's software product, from Simular Research, Inc. Published at simular.a
By Skyvern
Skyvern automates browser workflows with AI. No brittle selectors, no maintenance. Automate logins, form fills, data extraction, and more across any website. Skyvern is Skyvern's s
By TinyFish
TinyFish provides enterprise infrastructure for AI web agents. Run hundreds of websites simultaneously to navigate, authenticate, extract data, and automate web workflows with a fu
By UiPath
UiPath is an enterprise automation company selling robotic process automation, API integration and related AI features. Public marketing covers desktop and server digital workers f
Products
Claimed products appear first. Ranking packs and payment do not change this list.
- Agent S2· Simular
- Amazon Nova Act· Amazon Web Services
- ChatGPT Work· OpenAI
- Claude Computer Use· Anthropic
- Gemini 3.6 Flash· Google
- Induced· Induced
- Process Reasoning Engine· Automation Anywhere
- Skyvern· Skyvern
- TinyFish Web Agent· TinyFish
- UI-TARS Desktop· ByteDance
Also used in this category
These products have a different primary category so they do not compete for the same ranking queries. They are listed here because buyers still encounter them in this job.
Related categories
Relevant procurement and assurance guides
Frequently asked questions
What is Computer-use agents?
Agents that operate a desktop or OS GUI, not only a browser or API. Jobs: desktop gui automation with a model in the loop; os-level tool use; mixed app workflows on a vm.
What should not be listed as Computer-use agents?
Products whose buyer job is browser-only agents, IDE coding agents, or RPA with a scripted selector pack. Those belong on their own category page so search queries are not split.
Has AI Trustmark independently assessed every Computer-use agents supplier?
No. A category listing is descriptive. Independent assessment is shown only on company or product pages that carry Trustmark evidence.
What are Computer-use agents?
Agents that operate a desktop or OS GUI, not only a browser or API. Jobs: desktop gui automation with a model in the loop; os-level tool use; mixed app workflows on a vm.
What should buyers check when an AI agent can take actions?
Ask which tools, systems and write actions are allowed, whether those actions require human approval, and how they can be stopped. Category membership as an AI agent is a topic label, not permission to act in a buyer environment.
What logging should an AI agent provide?
Buyers should be able to see who the agent acted as, which tool was called, what data was sent, what changed, and when. Logs that omit write actions or store raw customer prompts without access control are incomplete evidence.
Should agent actions be reversible?
Where an agent can create, update, pay, message or delete, buyers should ask how a mistaken or injected action is stopped and reversed. If rollback is impossible, the residual risk and compensating controls should be explicit.
How is an AI agent different from workflow automation?
Workflow automation follows a designed path, even if a model fills a step. Agents choose steps toward a goal. Both can write to systems; the control model differs. Do not treat an RPA bot or a predefined workflow as an open-ended agent in procurement language.