Picovoice Launches Completely Free Usage Tier for Offline Voice Recognition — for Up to Three Users

Picovoice, creator of a software development kit for offline voice recognition and wake-word detection, has announced the launch of a free tier — which even allows for commercial use, for smaller projects.

Originally unveiled for the Raspberry Pi and other relatively high-performance processing platforms, Picovoice launched a variant of its software development kit earlier this year for resource-constrained devices — bringing secure offline fully-local wake world and speech-to-intent capabilities to microcontrollers, including the Arduino Nano 33 BLE Sense.

Now, that functionality is being made available to all free of charge — providing you're working on a project for no more than three active users per month.

"Anyone with an email address or a GitHub account can use Picovoice to train voice models and deploy them, even commercially, for free," the company explains. "This includes Porcupine wake word, Rhino Speech-to-Intent, and Cobra VAD [Voice Activity Detection] engines.

"No credit card is required. You don’t need to ever convert to a paying customer if you are working on a small project, personal or commercial. You can use all capabilities of Picovoice Console to train voice models and deploy them using any of our SDKs to your target device."

While Picovoice confirms that the free tier is open to commercial projects as well as hobbyists, the limit of three active users a month will bite the former pretty early on — though, at least, allows for early prototyping and development. The next-lowest pricing tier is $899 per month, which allows for up to 1,000 active users per month and offers email support to the free tier's reliance on community support via GitHub.

To support its free tier, which is open to users now, Picovoice has published documentation, provided access to a web-based console for project building, and has public repositories for its various offerings on GitHub.