Cursor voice chat, released on November 6, 2026, is a realtime spoken conversation with a Cursor agent. You talk, the voice agent answers out loud and hands coding work to the agent in your chat. It runs on Grok Voice and works in the Agents Window, the IDE, the web and Cursor for iOS.
Last updated: November 7, 2026
Table of contents
- What Cursor shipped on November 6, 2026
- How to start a voice chat
- What the voice agent can do during a call
- Voice chat compared with dictation
- Voice chat on Cursor for iOS
- Requirements, limits and privacy
- Frequently asked questions
- Sources
What Cursor shipped on November 6, 2026
The Cursor changelog entry “Voice Chat with Agents” introduced realtime voice calls with agents. You can describe a change out loud, ask what an agent is working on, and hear updates as the work moves forward.
The feature arrived one day after Cursor’s live app builder, covered in our Cursor Live Apps guide. It also extends the mobile workflow described in our Cursor Remote Control guide, because calls work in Cursor for iOS as well as on the desktop.
According to the voice chat documentation, Grok Voice handles listening and speaking, while your coding agent keeps the model you selected. With Privacy Mode on, voice chat runs with zero data retention.
For background on the speech model, see our Grok Voice 3.0 guide.
How to start a voice chat
Update Cursor first, then follow the steps from the official docs:
- Sign in to Cursor with your account. Sessions authenticated with an API key cannot start voice chat.
- Open an agent chat. In the Agents Window any chat works, including a new one. In the IDE, the chat needs at least one message.
- Wait until the agent is idle and clear the chat input. A waveform button appears where Send usually sits; while the agent runs, Stop takes that spot.
- Click the waveform button labeled Start voice chat. The microphone button next to it starts dictation instead.
- Allow microphone access if your computer, browser or phone asks. The call connects.
- Start talking. If you stay quiet, the agent checks in after a few seconds.
To change the voice, pick a default in the Voice or Agent Voice settings. You can also switch voices during a call; that choice sticks to the agent you are talking to.
What the voice agent can do during a call
Voice chat is speech-to-speech. According to the docs, when you ask for code changes, the voice agent hands the work to the coding agent in that chat and keeps talking with you while the work runs. During a call it can:
- Hand off work: send a task or follow-up to the coding agent in the current chat.
- Answer questions about your work: read diffs, runs and agents to tell you what changed and what is still running. These lookups are read-only.
- Search history: search earlier messages and calls with this agent.
- Switch Project workers: in a Project with workers, move the call to a different worker.
- Hang up: end the call when you say you are done.
The coding agent decides what is worth saying. It speaks short progress notes while it works, a brief answer when it finishes, and an update when workers it started report back. If it does not write a spoken update for a turn, the call reads its written reply instead. The call transcript appears in the chat.
Voice chat compared with dictation
Voice chat is separate from dictation, which earlier releases called Voice Mode. The prompting docs describe dictation as speech-to-text that does not talk back. The table below summarizes the official comparison.
| Voice chat | Dictation (formerly Voice Mode) | |
|---|---|---|
| Button | Waveform, Start voice chat | Microphone icon in the chat input |
| What you get | A spoken reply from the agent | Text in the chat input |
| Who sends the prompt | The voice agent hands off work | You review and send |
| Shortcut | None | Cmd+Shift+Space (Ctrl+Shift+Space) in the IDE; hold Ctrl+M in the Agents Window |
| Speech model | Grok Voice | Not stated in the docs |
Voice chat on Cursor for iOS
The Cursor for iOS docs list “Talk with agents” among the app’s features, including starting a voice chat from Siri. The app runs on iPhone with iOS 26.0 or later and iPad with iPadOS 26.0 or later, in English, and is available on Premium, Super, Ultra, Teams and Enterprise plans plus the legacy Start, Pro, Pro+ and Ultra plans.
- Siri: say “Hey Siri, call Cursor” to start a call with a new agent. Your iPhone needs to be unlocked.
- Live Activity: when iOS cannot show the call on the system call screen, the call appears on the Lock Screen and in the Dynamic Island. Mute and hang up from the Lock Screen or the expanded Dynamic Island; the compact Dynamic Island does not show those controls.
Cursor for iOS relies on Cloud Agents. If you are on Privacy Mode (Legacy), the mobile docs say to switch to Privacy Mode before using the app. Our Cursor Cloud Agents guide covers how those environments work.
Requirements, limits and privacy
- Account: a signed-in Cursor account. If you start a call while signed out or using an API key, it fails with “Realtime voice requires sign-in.”
- Microphone: if access was denied, Cursor shows “Microphone blocked. Allow it in your system or browser settings.” (on iOS, “Allow it in Settings”).
- Silence timeout: a call hangs up after about 10 minutes of silence from you while the agent has no work in progress.
- Daily limit: there is a daily limit on voice calls; Cursor tells you when you can call again. The number is not published.
- Hidden button: in the IDE, the button is also hidden for cloud agents that are still streaming and when dictation is turned off (see the FAQ for other cases).
- Privacy: with Privacy Mode on, voice chat runs with zero data retention, and cloud agent transcripts do not sync to other devices during a live call.
Cursor has not published separate pricing for voice chat. For plan costs and usage allowances, see Cursor usage limits on each plan.
Frequently asked questions
Does Cursor have voice chat?
Yes. Cursor added voice chat on November 6, 2026. It is a realtime, two-way spoken conversation with an agent in the Agents Window, the IDE, on the web and in Cursor for iOS. Start it with the waveform button labeled Start voice chat.
Is Cursor voice chat the same as Voice Mode?
No. Earlier releases used the name Voice Mode for dictation, which turns your speech into text in the chat input for you to review and send. Voice chat is a separate feature in which the agent replies out loud and the voice agent hands off work for you.
Which model powers Cursor voice chat?
Voice chat runs on Grok Voice, which handles listening and speaking. The coding agent in the chat keeps using the model you picked for it. You can choose among named voices such as Ara, Eve, Leo, Rex and Sal in Voice settings.
Why can’t I see the Start voice chat button?
The button appears only when the agent is idle and the chat input is empty; in the IDE the chat also needs at least one message. It is hidden in subagents, Project workers, archived chats, Claude Code chats and chats owned by another teammate.
Is there a limit on Cursor voice calls?
Yes. Cursor applies a daily limit on voice calls and tells you when you can call again if you reach it. A call also hangs up after about 10 minutes of silence from you while the agent has no work in progress. Cursor does not publish the daily number.
Does voice chat keep my audio?
Cursor states that with Privacy Mode on, voice chat runs with zero data retention. Call transcripts appear in the chat, and cloud agent transcripts sync to your other devices during a live call unless Privacy Mode is on.
Sources
- Cursor changelog — Voice Chat with Agents (November 6, 2026)
- Cursor Docs — Voice chat
- Cursor Docs — Prompting agents (Voice input)
- Cursor Docs — Cursor for iOS