# Tovera

- **Website:** <https://heywillow.io>
- **Primary alias:** `heywillow.io`

## Description

Willow is a practical and open-source platform designed for voice assistants and other applications. Developed using the ESP IDF technology and primarily targeting the ESP32-S3-BOX hardware family from Espressif, Willow aims to provide competitive performance, accuracy, cost-effectiveness, and functionality similar to Amazon Echo/Google Home. With its 100% open-source nature, Willow can be completely self-hosted by the user, making it an affordable choice with readily available commercially available hardware. One of its standout features is its exceptional speed, with response times faster than Alexa/Echo or Google Home, completing actions in 500ms or less. Willow also boasts high wake word accuracy and low false activation, thanks to its Willow Inference Server or device-based command recognition. It has been extensively tested for reliability, with a failure rate of less than 1%. Additionally, Willow offers flexibility, allowing users to choose between using the Willow Inference Server or performing command recognition on the device itself. The platform is fully customizable and supports multiple wake words. Furthermore, Willow prioritizes privacy, offering users complete control over their data by enabling them to check the source, build and flash the platform themselves, and even proxy through another server for traffic inspection. Users can utilize Willow on their own servers, solely with local commands, or on a network without internet access. The Willow platform is highly practical and aesthetically pleasing, designed to seamlessly blend into any home or office environment. With an approximate hardware cost of $50 (plus a USB-C power supply), Willow is an affordable and energy-efficient solution, consuming only 100mW. The current range of supported features includes the Wake Word Engine for easy activation, Voice Activity Detection for seamless communication, and support for popular platforms like Home Assistant, openHAB, and generic REST APIs. Willow also offers great far-field performance, delivering accurate wake and speech recognition even from up to 25 feet away in challenging environments. Additionally, it boasts excellent audio quality, with features such as automatic gain control, acoustic echo cancellation, noise suppression, and blind source separation. For users in Wi-Fi-congested areas, Willow provides the option of audio compression to reduce airtime on the 2.4 GHz frequency. Moreover, Willow supports the ESP32-S3-BOX-3 with a color LCD and capacitive multi-point touchscreen, along with an initial user interface for an enhanced user experience. Users can configure up to 400 commands on the device itself or choose to self-host the Willow Inference Server for advanced speech transcription. Overall, Willow offers a versatile and cost-effective solution for voice assistants and other applications, providing users with the opportunity to customize and control their voice interactions while ensuring privacy and ease of use.

## Industries

- 🖥 Computers Electronics and Technology
- 🖥 Smart Homes and Home Automation _(under Computers Electronics and Technology)_
- 🖥 Artificial Intelligence and Machine Learning _(under Computers Electronics and Technology)_

## Links

- [twitter](https://twitter.com/toverainc)
- [github](https://github.com/HeyWillow)

## Logos & icons

- logo _(primary)_ — PNG, SVG — [download](https://cdn.brandfetch.io/idE1snTON3/w/820/h/925/theme/dark/logo.png?c=1bxid64Mup7aczewSAYMX&t=1749715614627)
- icon _(primary)_ — JPEG — [download](https://cdn.brandfetch.io/idE1snTON3/w/400/h/400/theme/dark/icon.jpeg?c=1bxid64Mup7aczewSAYMX&t=1749715614703)

## Colors

| Name | Hex | Theme |
| --- | --- | --- |
| Bleached Cedar | #271928 | dark |
| White | #ffffff | light |

## Fonts

- var(--md-text-font-family) — asset _(custom)_

---

_Source: <https://brandfetch.com/heywillow.io>_
_See [/llms.txt](/llms.txt) for a full list of agent-readable pages, and [/auth.md](/auth.md) for how agents authenticate._