Embed the world’s #1 On-Device Voice AI Technologies

Krisp’s On-Device Voice AI SDKs are integrated into more than 150 million devices powering mobile, desktop and browser apps in audio conferencing, podcasting, streaming, and recording.
The SDKs include real-time Background Noise and Voice Cancellation, Accent Localization and Speech-to-text, all running on-device.

We power the world’s leading voice and video communication platforms

Great customer experience starts with clear communication

10%

Decrease in
AHT

8%

Increase in
CSAT

30%

Increase in
cost savings

On-Device Voice AI technologies

Krisp is the leading developer of advanced voice processing technologies. Our award winning, AI-powered voice technologies leverage deep neural networks to achieve an unprecedented level of voice quality enhancements that change the way business and communication is done.

Krisp SDKs are integrated directly into an application’s audio path. All voice processing happens on-device and voice is never processed or stored in the cloud. Krisp SDKs have no server connectivity.

Krisp SDK for Developers
microphone Background Noise Cancellation

Model designed for microphone (outbound/uplink) audio stream. Noise robust for use in contact centers, offices and home environments. Includes de-reverberation to remove room echo. Physical microphone and language independent.

room echo Background Voice Cancellation

Model designed for microphone (outbound/uplink) audio stream to remove other voices near the primary speaker wearing a headset or ear buds. Fixes the contact center cross-talk problem, as well as blocking competing voices in office and at home. No enrollment or training of the user’s voice is required. Background Noise Cancellation included. Language independent.

Voice activity detector Inbound Noise Cancellation

Model designed for the speaker (inbound/downlink) audio stream. Noise robust for CCaaS and UCaaS platforms so that customers calling in on mobile or PTSN have their background noise removed so that agents can have clear communications bi-directionally. Physical speaker or headset independent. Language independent. Intelligent pass-through of ringtones.

Voice activity detector Noise Cancellation for Improved STT

Model designed for the inbound audio stream. Integrated on-server in front of 3rd party and proprietary STT platforms. Removes background noises for improved Word Error Rate (WER) and reduced hallucinations with minimal latency impact. Language and accent independent.

Voice activity detector Accent Localization

Model designed for changing the accent of a contact center agent in real-time from their local accent to that of the customer, improving customer experience while providing a significant ROI. The first model availability is for India-based agents serving North American customers in English. Local accent robust, gender independent, noise robust and microphone/headset independent.

Voice activity detector On-Device Transcriptions (STT)

On-device Speech-to-Text (STT) model designed to transcribe users’ online communications in English in near-real time. Krisp STT is noise and accent robust, translating to top performing WER. No internet connection required, making transcriptions available to applications and devices independent of connectivity. On-device STT also translates to much lower cost, by removing transmission and server costs associated with cloud-based solutions. Available for Win and Mac OS.

Krisp Noise Cancelling app demo

Play and toggle Krisp on and off
play button
Without Krisp With Krisp
Try different noise type

Accent Localization

Supercharge your call center with AI Accent Localization

Runs on end user device

Krisp technology is designed and optimized to run on the end user’s device. This allows you to:

Save money on server infrastructure

No need to invest in server processing since everything with Krisp is done right on the user’s machine.

Eliminate the need to implement server-side code

Integrating our technology is seamless and easy since servers are taken out of the equation.

Avoid sending user audio and data to external servers

Processing is done faster and more securely without the added step of sending data to an external server.

Supported platforms

Krisp’s Audio SDK library is available for Windows, Mac, Linux, Web (JS/WASM), iOS and Android platforms.

arm
intel
amd
c++ Compiled C library
Krisp runs on Windows, Mac, Linux, iOS and Android and can be compiled for special-purpose embedded systems.
Javascript Logo Javascript library
Krisp’s SDK easily integrates into your web application. Requires WASM support (Chrome, Edge, Firefox).

Organizations worldwide love Krisp

Krisp config in discord app

Leverage Krisp’s brand

Krisp Voice SDK brand can be highlighted in your application, assuring your users that they have the world’s best voice processing technologies as part of their everyday communication experience.

Contact us today

Tell us about your solution, and we’ll tell you how Krisp can help

Complete the form at the link below to tell us about your product and use case. We will contact you to discuss how to get started with integrating Krisp SDKs.