88Z AI, 27 August — Google releases Gemini 3.5 Transcribe, which automatically removes filler words and self-corrections. It is about 70% faster than Chirp 3, with a real-time error rate of 5.5%. The model supports 85 languages and multi-speaker audio, and is already used in Pixel 11 Gboard.
At a glance
- Speed increased by about 70% over predecessor, with real-time error rate down to 5.5%
- Supports 85 languages and three-speaker audio, already used in Pixel 11
The update
Google has officially launched the Gemini 3.5 Transcribe speech-to-text model. This model can automatically identify and remove filler words such as "um" and "uh," as well as self-corrections during speech, thereby generating smoother and more fluent text. In terms of performance, it is approximately 70% faster than the previous Chirp 3 model, while reducing the real-time speech recognition error rate to 5.5%. Currently, the model supports recognition for up to 85 languages and can process pre-recorded audio files with up to three speakers. This new technology has already been implemented and integrated into the "Rambler" feature of the Gboard keyboard on the Pixel 11 smartphone, providing users with a more efficient voice input experience.
Sources & editorial note
This article is an AI-assisted synthesis by 88Z AI based on the public sources below. Product claims and availability are attributed to their publishers. It does not represent independent testing or first-hand reporting.
- IT之家Source published
Source dates above refer to the original announcements; the date under the headline is this article’s publication on 88Z AI.