# Is Wispr Flow safe? What it keeps, and what to use instead

Canonical: https://vibetyper.com/blog/is-wispr-flow-safe
Description: Wispr Flow's help center says training and cloud storage start on. What it keeps, and an app that keeps no recordings on its servers. Free to start, no card.
Published: 2026-11-05T00:00:00.000Z
Updated: 2026-11-05T00:00:00.000Z
Tags: wisprflow, privacy, comparison, alternatives, voice-typing

If you're weighing up Wispr Flow and want to know what happens to what you say, this is the short version. I make Vibe Typer, a dictation app that competes with Wispr, so keep that in mind. That's why everything I say about Wispr here is quoted from Wispr's own pages and linked, so you can check each line yourself. I read every page on 5 Nov 2026.

## What Wispr Flow does with your voice

Wispr Flow turns your speech into text in the cloud. Its [data controls page](https://wisprflow.ai/data-controls), checked 5 Nov 2026, says "Transcription always occurs on the cloud." Its [security overview](https://docs.wisprflow.ai/articles/3467817258-security-and-compliance-faq), checked 5 Nov 2026, says it "processes and stores customer data in the United States."

What happens after that comes down to two switches.

## Two switches that start on

Wispr has two data switches under Settings → Data and Privacy. Its [help article on data sharing and cloud storage](https://docs.wisprflow.ai/articles/9609615338-private-cloud-sync-and-data-sharing-preferences-in-wispr-flow), checked 5 Nov 2026, describes them:

- **Improve the model for everyone:** "Allows or blocks use of your Notetaker and dictation content to improve Wispr's models."
- **Dictation cloud storage:** "Stores dictation transcripts, audio and history on Wispr's servers for cross-device access."

The same article says "Both switches default to on unless restricted". So out of the box, Wispr keeps your transcripts, audio and history on its servers, and can use what you dictate to improve its models.

Wispr's [privacy page](https://wisprflow.ai/privacy), checked 5 Nov 2026, says "Your audio is deleted as soon as it's transcribed, unless you've turned on Dictation cloud storage." Going by the help article, that switch starts on.

Wispr adds two things. Its security overview says training is "on by default for Free and Pro, with the choice shown during onboarding, and you can turn it off". And the help article says "Enterprise model training is Off for everyone, with no enable switch".

Two more if you do switch them off. "Signing out resets the cloud-storage preference to on", so check it again after you sign back in. And "A dictation you choose to report is retained by Wispr and can be used to train and evaluate its models."

## What still goes to Wispr with training off

The help article is clear on this: "Turning it off does not stop cloud processing." Your audio and transcripts are held back from model work, but "Flow still uploads dictation details such as word count, duration, app, URL, timestamp, app version, microphone, detected language and edit statistics."

The data controls page adds that "Wispr may collect usage statistics such as the number of words you have dictated, regardless of your data controls." The help article also says "Wispr does not sell your data; aggregate usage statistics, such as words dictated, do not include dictation contents."

## What Context Awareness reads from your screen

Context Awareness is a third setting. Its [help article](https://docs.wisprflow.ai/articles/4678293671-feature-context-awareness), checked 5 Nov 2026, says "It's on by default." While it's on, "context can accompany your dictation: app info, textbox contents (before, selected, and after the cursor), on-screen text, variable and file names in coding apps, your user identifier within the app, and the apps in your current session."

On screenshots, Wispr says "Flow works from text and does not take screenshots." It also says "Captured text, extracted names, and counts for a dictation are cleared when that dictation ends", "Password field contents are never read", and Flow "skips context reading entirely in banking and financial apps".

Turning it off doesn't stop everything: "On Mac and Windows, Flow still sends the app name and website domain with each dictation to set the right tone." On iPhone, "The keyboard sends the text before and after your cursor, any selected text, and the app name with each dictation. There's no Context Awareness switch on iPhone."

Wispr's own note: "For sensitive environments (legal, healthcare, or regulated data), turn Context Awareness off in Settings → Data and Privacy."

## Meetings and Notetaker

Cloud storage doesn't cover Wispr's meeting notes. The data sharing article says "Notetaker transcripts and summaries are stored in Wispr's cloud, with no setting to opt out." The data controls page says "You can configure notetaker transcript retention in Settings > Data and Privacy."

For the recordings themselves, the security overview says "Meeting audio uploaded to Wispr is kept for 7 days from upload to support troubleshooting and speaker attribution, then deleted; transcripts and summaries remain."

## Security and certifications

From the security overview:

- "Wispr encrypts service data in transit and at rest. It is not end-to-end encrypted in the strict sense, because the service must process audio to produce a transcript."
- Its Trust Center holds "Wispr's SOC 2 Type II report, the ISO/IEC 27001:2022 certificate".
- "Meeting Notetaker launched after those audit periods and was not assessed in the SOC 2 Type I, the SOC 2 Type II, or the ISO 27001 certification audit."
- "HIPAA-ready dictation requires a signed Business Associate Agreement (BAA)." The privacy page says you can "Sign a Business Associate Agreement to enable HIPAA support for dictation, on any plan."

On the AI companies it uses, the data controls page says "Wispr always maintains zero data retention agreements with all third-party AI providers" and "no third party AI providers can use your data for model training." It adds that "Certain features, such as briefs for Notetaker, rely on features from third party AI providers that are not supported under zero data retention."

## So, is Wispr Flow safe?

That's your call, and it depends on what you dictate. Here's what Wispr's own pages say, in one place:

- Your voice is always turned into text in the cloud, and your data is processed and stored in the United States.
- Training and cloud storage both start on for Free and Pro. You can turn both off.
- Context Awareness starts on and can send text from your screen with a dictation. It doesn't take screenshots.
- With training off, dictation details such as the app, URL, word count and duration still go to Wispr.
- Meeting transcripts and summaries stay in Wispr's cloud.
- Wispr holds SOC 2 Type II and ISO 27001, and offers a BAA for HIPAA dictation.

If you'd rather your recordings and transcripts weren't kept at all, with nothing to switch off, that's Vibe Typer.

## What to use instead: Vibe Typer

Vibe Typer keeps none of your recordings or transcripts. It's cloud too: your voice is turned into text on servers. What's different is what happens after, and there's no switch to find. By design:

- Your audio is processed in memory and never stored on Vibe Typer's servers.
- Your text comes straight back to your device and isn't stored on the servers either. If your connection drops, your device sends the audio again, so the servers don't keep a copy for retries.
- The speech and AI services I use don't keep your recordings or your text, and nothing you say is used to train AI models, mine included.
- A dictation sends your audio and nothing from your screen: no screenshots, no app name, no window title. When you ask Vibe Typer to rewrite or reply, it sends the text you selected or copied, because that's the text it's working on.
- Your history stays on your computer and is never uploaded. Disable history clears it and stops saving new takes. Here's [how your history stays on your computer](/docs/local-history-and-privacy).

What Vibe Typer does keep is short: your email address, your settings so they follow you between devices, counts like audio minutes and words, and your IP address and request details in sign-in records and server logs, for security. You can delete your account in the app under Settings → Account. The full list is in the [privacy policy](/privacy).

It's one app on Apple Silicon Macs, Windows, Linux and iPhone, with one account across all of them.

## Wispr Flow and Vibe Typer side by side

| | Wispr Flow, out of the box | Vibe Typer |
| --- | --- | --- |
| Where speech becomes text | Cloud | Cloud, processed in memory and never stored |
| Audio of your dictations | Kept on Wispr's servers while Dictation cloud storage is on, and it starts on | Never stored on the servers |
| Transcripts and history | Kept on Wispr's servers while cloud storage is on | Never stored on the servers. History stays on your computer |
| Used to improve models | On by default for Free and Pro | No |
| Text from your screen sent with a dictation | Context Awareness starts on: text around your cursor, on-screen text and app info | None |
| Screenshots | None, Wispr says | None |
| Switches to change for the most private setup | Three: training, cloud storage and Context Awareness | None |

For everything else, from Linux support to price, see the full [Wispr Flow vs Vibe Typer comparison](/blog/wispr-flow-vs-vibe-typer) and the [best Wispr Flow alternatives](/blog/wispr-flow-alternatives).

## Get started

If you want a dictation app with nothing to switch off first, [download Vibe Typer](/downloads). Press your hotkey, talk, and clean text lands in whatever app you're in. Setting up on Linux? Here's [how to set up a Wispr Flow alternative on Linux](/blog/wisprflow-alternative-linux-guide). On Windows, here's the [Windows guide](/blog/wisprflow-alternative-windows-guide). Free to start, no card.

## FAQ

### Is Wispr Flow safe?

That depends on what you dictate and which switches you change. Wispr Flow holds a SOC 2 Type II report and an ISO 27001 certificate, and says it doesn't sell your data. Its help center also says model training and cloud storage both start on, so it keeps your audio, transcripts and history on its servers unless you switch cloud storage off. If you'd rather none of that were kept, [Vibe Typer](/downloads) keeps no recordings or transcripts on its servers, by design.

### Does Wispr Flow store your data?

By default, yes. Wispr's help center says Dictation cloud storage keeps your transcripts, audio and history on its servers and starts on. Turn it off and Wispr says audio is discarded once it's transcribed, though it may still collect usage statistics such as words dictated regardless of your data controls. Vibe Typer stores no recordings or transcripts on its servers at all.

### Does Wispr Flow train on your voice?

It can. Improve the model for everyone is on by default for Free and Pro, shown during onboarding, and you can turn it off. It's off for Enterprise. A dictation you report can be kept and used for training either way. Vibe Typer doesn't use your voice or your text to train models.

### What is Wispr Flow Privacy Mode?

It's the old name for Wispr's model training setting, now called Improve the model for everyone. Turn that off and your dictations stay out of training, though Wispr says it doesn't stop cloud processing. Vibe Typer doesn't train on what you say and keeps no recordings or transcripts on its servers, so there's no switch to hunt for.

### Does Wispr Flow take screenshots?

Wispr says it works from text and does not take screenshots. With Context Awareness on, which is the default, a dictation can carry text from your screen: the text around your cursor, on-screen text and app info. Vibe Typer takes no screenshots, and a dictation sends nothing from your screen.

### Is Wispr Flow HIPAA compliant?

Wispr says HIPAA-ready dictation requires a signed Business Associate Agreement, and its privacy page says you can sign one on any plan. Its help center says the Meeting Notetaker wasn't assessed in its SOC 2 or ISO 27001 audits.

### Is Wispr Flow secure?

Wispr encrypts data in transit and at rest, and its Trust Center holds a SOC 2 Type II report and an ISO 27001 certificate. It says it isn't end-to-end encrypted in the strict sense, because it has to process your audio to transcribe it, and it processes and stores data in the United States. Vibe Typer takes a simpler route: no recordings or transcripts are kept on its servers, so there are none there to protect.

### What is a more private alternative to Wispr Flow?

[Vibe Typer](/downloads). It's cloud too, but your audio is processed in memory and never stored on its servers, and no transcripts are kept there. There's nothing to switch off first. It's one app on Apple Silicon Macs, Windows, Linux and iPhone. Free to start, no card.

### Does Vibe Typer keep my recordings?

Never on its servers. Audio is processed in memory and not stored, and transcripts aren't stored there either. On your computer, recordings are kept only if you turn on Save recordings, which starts off. Disable history clears your history and stops saving new ones.