Voice
Talk naturally
Hold push-to-talk to dictate, or turn on Voice mode to say "Hey Yoo", hear spoken replies, ask follow-ups, and interrupt when needed.
- Speech-to-text
- Spoken responses
- Stop and New Chat controls
The AI assistant that helps you do computer tasks.
Talk, browse, organize files, read documents, and handle visible desktop steps while sensitive actions stay under your control.
What it can help with
ActionWright is useful when a task is easy to explain but annoying to do manually.
Voice
Hold push-to-talk to dictate, or turn on Voice mode to say "Hey Yoo", hear spoken replies, ask follow-ups, and interrupt when needed.
Task window
Type longer requests, attach files, choose a folder, open the agent browser to watch it work, and answer follow-up questions in one place.
Browser
Search, read pages, compare information, click, type, scroll, and work through browser flows with approval checkpoints.
Files
Find files, summarize documents, create folders, edit text, rename, copy, move, and organize local work.
Desktop
When a task needs the screen, ActionWright can help click, type, paste, scroll, focus windows, and open apps.
Providers
Connect supported providers for text and voice. Keys are validated before saving and stored in the OS credential vault.
Common requests
Use ActionWright for research, document work, browser steps, and local cleanup without writing a formal prompt.
Example
“Summarize these PDFs and pull out decisions, risks, and next steps.”
A clean brief from files you attach or choose from a folder.
Example
“Compare these options online and give me a short recommendation.”
Web research with useful notes and page context.
Example
“Clean up my Downloads folder by grouping files I should keep.”
A preview you approve before files are moved or cleaned up.
Example
“Help me fill this form, but pause before submitting it.”
Visible browser assistance with confirmation at sensitive steps.
User control
ActionWright shows progress as it works and separates routine steps from actions that need your confirmation.
Read the safety docsTell ActionWright what you want done by voice or text.
Add context if you want: attached files or a selected folder.
Review sensitive actions, then stop, take over, or continue with follow-up instructions.
Works with
ActionWright connects to AI providers you configure and works in the computer contexts you approve.
ActionWright is a Windows app that helps with the browser, files, and visible desktop actions.
OpenAI, Gemini, Grok, and OpenRouter include voice. Anthropic has no voice, so add one of the others as a separate voice provider for speech.
The browser agent always uses its own Chrome profile. Sign in there once and it remembers, so ActionWright can pause for you and continue.
FAQ
Question
ActionWright can help with voice and text tasks, websites, selected-folder file work, and visible desktop actions.
Question
No, it acts only on tasks you give it. With Computer Use on (the default), it uses your mouse and keyboard during a task, asks before sensitive actions, and stops when you press ctrl+esc.
Question
Sometimes. Recycle Bin moves always ask, and payments are built to always ask. Sends, form submissions, deletions, and account changes ask unless your own request asked for them. File changes ask unless you chose Always Allow.