[ embedded voice AI powered by deep learning ]

Picovoice provides core technology to embed privacy-protecting voice AI into any product. It enables companies to build products that can be activated and controlled using voice without cloud connection.



Runs locally without an internet connection. Protects user's data as nothing is sent to the cloud.



Uses proprietary deep learning technology that enables cutting-edge models to run on commodity (embedded) hardware efficiently.



Resilient to noise, reverberation, and works across a variety of accents. Works everywhere for anyone.

plug and play

Plug and Play

Enables customizations within seconds. Reduces time to market while promoting user's identity.

cross platform


Runs everywhere. Picovoice’s core technology is OS/CPU independent. Linux, Mac, Windows, Android, iOS, watchOS, Raspberry Pi, ARM Cortex-A, ARM Cortex-M, and an ever-growing list of embedded platforms are supported.

open technology

Open Technology

Picovoice empowers users to evaluate its technology independently. Facilitates data-driven decision making.

Wake Word Engine (Voice Control Engine)

The wake word detection SDK enables makers to build always-listening products that can be activated and controlled using voice. It makes it possible to activate the device similar to “Alexa” or “OK Google” but using your hotword of choice promoting your identity, brand, and product. Additionally, the library allows to control the device via configurable voice commands.

zero lead time

Zero Lead Time

Uses proprietary AI algorithms to build models for any custom wake word/command within seconds. Removes the hassle of time-consuming and costly data gathering phase.



Can detect many commands/hotwords concurrently with virtually no additional CPU/memory footprint.



Can run with as low as 240 KB of memory and 3.4% CPU usage on a Raspberry Pi 3. Natural fit for IoT applications. Runs on Android, iOS, watchOS, Raspberry Pi, ARM Cortex-A, ARM Cortex-M, etc.



Outperforms existing solutions with high margins in clean and noisy environments.



Has a fast-growing and vibrant open-source community. Expedite your development by reusing numerous publicly-available references designs.

language independent

Language Independent

Can create models for almost any phrase in any language.

Speech to Text Engine

The speech to text SDK enables building conversational interfaces without the need to any cloud connection. It is powered by a novel end-to-end learning algorithm which makes it possible to perform accurate speech recognition on IoT platforms with extremely limited memory/CPU budget.

open vocabulary

Open Vocabulary

No limit on the size of vocabulary. Provides large vocabulary transcription capabilities on embedded platforms.


Continuous and Real-Time

Transcribes audio in real time. Reduces latency and improves the user experience.



Runs in real-time with only 5.6 MB of memory and 25% CPU usage on a Raspberry Pi 3. Natural fit for IoT platforms. Android, iOS, Raspberry Pi, and a growing number of IoT platforms are supported.


Picovoice is a team of applied scientists and engineers who strive to build a future where our lives are enhanced with ambient voice AIs that respect our privacy. Picovoice is founded by Alireza Kenarsari. Prior to Picovoice Alireza was a Senior Engineer in Amazon and has been an early engineer in a few successful technology startups (one reached IPO). He is the inventor of five US patents within the fields of deep learning and speech recognition. Read more about the beginnings of Picovoice here.


[July 23, 2018] Picovoice ranks among top 10 open source machine learning projects. We are extremely excited to be on this list alongside names such as Facebook Research, Salesforce, Baidu Research, and NVIDIA. Read more here.