GatwAI

Krisp

Evaluate Krisp for clearer calls and meeting notes, with current plans, local audio filtering, cloud processing and practical evaluation workflows.

Provider sources checked

What does Krisp do?

Krisp starts during the conversation itself: it filters distracting noise and adds transcription, recording and summaries. The current meeting assistant also offers accent features; call-centre solutions and developer tools are separate product areas. For ordinary teams, the key decision is whether the immediate priority is better intelligibility or better meeting documentation. Its integration as an audio tool distinguishes Krisp from services that only join as a visible meeting bot. This can simplify some workflows, but correct microphone and speaker selection matters. A pilot should use the actual headset: filtering may reduce unwanted noise while also affecting parts of a quiet or highly variable voice.
Editorial research · 2026-11-14How we research →

Who is this tool for?

  • Remote teams working in varied acoustic environments.
  • Customer-support teams with recurring calling and documentation needs.

Strengths

  • Audio enhancement during the conversation.
  • Meeting notes without an additional joining bot.
  • Product areas for individuals, teams and integration.

Limitations and requirements

  • Audio quality depends on hardware, room and filter settings.
  • Bot-free capture still requires clear recording arrangements.
  • Cloud notes involve a different data flow from noise filtering alone.

Who is behind the product?

Company
Krisp Technologies, Inc.
Headquarters
Berkeley, California, US, according to the website contact address.
Company founded
2017 according to an official company announcement.
Product launch
The app's original launch date was not separately established here.

Expanded from speech enhancement into meeting AI, call-centre functions and developer components.

Suggested ways to use it

Suggested workflows or verification steps. These are not hands-on tests; for unverified products they are not a recommendation to use the service.

Suggested: pilot a headset configuration

  1. Record an agreed sample conversation with typical keyboard and room noise.
  2. Toggle Krisp and have another person assess clarity and possible voice distortion.
  3. Document the appropriate device configuration before rolling it out to the team.

Goal: Audio settings matched to the actual hardware.

Suggested: document a conversation

  1. Agree recording and storage beforehand; choose a non-sensitive meeting.
  2. Generate notes and compare tasks with the conversation.
  3. Correct names, dates and owners; share only the necessary report and set retention.

Goal: Checked meeting notes with a limited audience.

Pricing, allowances and data

What should you check before choosing a plan?

Meeting AI offers a seven-day trial without a credit card. Core costs US$16 per user/month, or US$8 equivalent with annual billing; Advanced costs US$30 or US$15. Listed storage is 10 GB and 60 GB respectively. Enterprise is quoted separately, as are additional call-centre and API requirements.

What should you check about data and permissions?

For noise cancellation alone, the privacy policy says audiovisual data does not leave the device. The meeting assistant may instead store transcripts, recordings and summaries and send content to processing providers. Non-English transcription can involve temporary audio that is deleted after transcription. Verify the selected mode and retention before use.

External reviews and available evidence

Zapier's 6 April 2026 hands-on review describes effective filtering and possible voice distortion. The integration vendor discloses commission links.

Reported positives

  • Specific device-setup observations.

Reported cautions

  • Hardware-dependent; no GatwAI listening test.

These reports reflect the linked authors’ experiences. Test date, plan and task affect results; they do not establish a representative overall rating.

Product development and notable changes

  • Current meeting-AI packages

    Core, Advanced and Enterprise plus VIVA 2.5 are currently advertised; review date, not release date.

    Source for this change ↗

Which alternative fits which task?

tl;dv →

tl;dv when searching and analysing a meeting library is the priority.

Granola →

Compare Granola for combining personal notes with AI assistance.

Frequently asked questions

Is Krisp entirely offline?

Not universally. Noise filtering alone and the meeting assistant use different data flows.

Does bot-free mean no recording?

No. The absence of a visible bot says nothing about whether audio is captured or transcribed.