Official documentation
Speech-to-Phrase brings voice home (Voice chapter 9)
Home Assistant project, Home Assistant blog, 2025
Read the originalRead on October 17, 2026
- What it is
- The February 2025 Home Assistant blog post that introduced Speech-to-Phrase, with its speeds on a Raspberry Pi 4 and Raspberry Pi 5, its launch languages and the new prefer-local setting for LLM agents.
- How much weight it carries
- Written by the team that builds the software, so it is reliable on what the project built and why. The speed figures are the project's own and were measured on their test setups, not independently.
Facts taken from this source
- Home Assistant says that before it added the prefer-local setting, many easy-to-run commands were being sent to a language model, which slowed responses and wasted tokens.
“we noticed many easy-to-run commands were being sent to an LLM, which can slow down things and waste tokens.” (Speech-to-Phrase introduction)
- Home Assistant's opt-in analytics show that over 50 percent of Home Assistant OS users run their homes on affordable, low-powered machines such as the Home Assistant Green or a Raspberry Pi.
“Our opt-in analytics shows over 50% of the Home Assistant OS users are running their homes on affordable, low-powered machines like the Home Assistant Green or a Raspberry Pi.” (Speech-to-Phrase introduction)
- At its launch, Speech-to-Phrase supported English, French, German, Dutch, Spanish and Italian, which Home Assistant said covers nearly 70 percent of its users.
“Speech-to-Phrase is launching with support for English, French, German, Dutch, Spanish, and Italian ... covering nearly 70% of Home Assistant users.” (Speech-to-Phrase introduction)
- Home Assistant reports that the Raspberry Pi 5 processes Speech-to-Phrase commands seven times faster than the Raspberry Pi 4, at 150 milliseconds per command.
“The Raspberry Pi 5 processes commands seven times faster, clocking in at 150 milliseconds per command!” (Speech-to-Phrase introduction)
- In February 2025, Home Assistant said that Whisper on a Raspberry Pi 4 takes at least 5 seconds to turn speech into text, which is why it recommended at least an Intel N100 for fully local voice.
“takes at least 5 seconds to turn your speech into text” (Speech-to-Phrase introduction)