|

ChatGPT Can Now Do Work by Voice on Your Phone: What’s New and How to Use It

AI Tools

On September 23, OpenAI brought voice-driven agent features to the ChatGPT mobile app. You can now ask ChatGPT out loud to draft a document, write an email, summarize your Slack messages, or build a presentation — and it does the work, not just the talking. Here’s what’s included, who gets it, and how to use it well.

What’s in this guide

  1. What’s new
  2. What you get on each plan
  3. Practical ways to use it
  4. Tips for better results
  5. Privacy and limits
  6. FAQ

What’s New

Until now, ChatGPT’s voice mode was mostly for conversation. The update connects voice to ChatGPT’s agent tools on mobile, so spoken requests can trigger real workflows. It builds on July’s desktop release of GPT-Live, OpenAI’s model for more natural live conversation, which was added to the desktop Work and Codex tabs.

  • Create documents and draft emails by voice
  • Summarize Slack messages
  • Build websites and presentations
  • Use a cloud browser to complete web tasks
  • Switch between voice and text in the same conversation
  • Start on your phone and continue on desktop
  • Richer text output alongside spoken answers

What You Get on Each Plan

Plan Voice work features
Plus and Pro Full Work tab: agent tasks, documents, emails, presentations, websites, cloud browser, finance features
Free and Go Plugins and connected apps only

Business and Enterprise availability can depend on workspace admin settings. Features may roll out gradually, so update the app if you don’t see them yet.

Practical Ways to Use It

On your commute

“Summarize what I missed in Slack this morning and list anything that needs my reply.”

After a meeting

“Turn what I’m about to say into a follow-up email with action items and owners.”

Idea capture

“I’m going to talk through a blog post idea. Turn it into an outline with headings.”

Quick decks

“Make a five-slide presentation summarizing our Q3 results for the team meeting.”

Tips for Better Results

  • Say the output format up front: “a short email,” “a one-page doc,” “five slides.”
  • Give context in one go: who it’s for, the tone, and the deadline. Voice makes it easy to ramble — structure helps.
  • Review before sending: treat voice-created drafts as drafts. Check names, numbers, and dates before anything leaves your account.
  • Finish on desktop: use mobile to start and capture, then polish formatting on a bigger screen.
  • Ask it to push back: if you want a real critique, say so — AI assistants tend to agree with you by default.

Privacy and Limits

Summarizing Slack or drafting emails means giving ChatGPT access to those accounts. Connect only the apps you actually need, and check your workspace’s AI policy before connecting work tools. Be careful using voice in public spaces, where others may hear sensitive details. And remember that agent features can make mistakes: they are best for drafting and summarizing, with you making the final call.

FAQ

Is this available on the free plan?

Partly. Free and Go users get plugins and connected apps; the full Work tab features require Plus or Pro.

Can I start by voice and finish typing?

Yes. You can switch between voice and text mid-conversation and continue on desktop.

Does it send emails automatically?

Treat everything it creates as a draft and review it before sending. Always check what you’re approving.

Related Reading on FutureLume

The Bottom Line

Voice-driven agent features turn dead time — walking, commuting, waiting — into time you can use to get drafts and summaries done. Plus and Pro users get the most out of it. Use it to start work quickly, and keep yourself in the loop before anything goes out.

Similar Posts

Leave a Reply

Your email address will not be published. Required fields are marked *