Fact · Dictation software
In a 2022 post, the whisper.cpp maintainer said the tiny.en Whisper model can run the real-time examples on a Raspberry Pi 4 Model B using 3 CPU threads.
Checked against its source on October 17, 2026
In the source's own words
“This allows to run the above examples on a Raspberry Pi 4 Model B (2018) on 3 CPU threads using the tiny.en Whisper model.”
Real-time transcription on Raspberry Pi 4 (whisper.cpp discussion #166), Georgi Gerganov, GitHub, ggml-org/whisper.cpp, 2022, Maintainer's post, More information section.
Open the original- What it measures
- The examples use a streaming mode with a 4 or 7.5 second step. To speed things up, the post cuts the encoder context from 1500 to 512 with the -ac 512 flag, and the microphone is a USB camera's.
- What it does not show
- A maintainer's demonstration, not a timed benchmark. It uses the English-only tiny model, a reduced encoder context and a poor microphone, so expect different results with other models, settings and audio.