Google announced a speech-to-text model for Gboard, Gemini on macOS, developer tools and eventually Chrome, emphasizing speed, multilingual support and automatic cleanup. Gemini 3.5 Transcribe removes verbal fillers and corrections, but its polished output can differ from the speaker's exact words. Google calls the model Gemini 3.5 Transcribe. The company reported about seventy percent faster processing than Chirp 3.
Google reported a 5.5 percent live-speech error rate in its tests. The model supports eighty-five languages and up to three speakers in prerecorded audio. It can remove fillers and revise text when a speaker corrects a phrase. These reported details define the immediate scale, timing and institutions involved without treating preliminary statements as final findings.
The model already powers the Rambler feature on Pixel 11 devices. Access is expanding through Gemini on macOS, AI Studio, Antigravity and the Gemini API. Google said Chrome support would follow later. The chronology separates the new development from older background and keeps official descriptions attached to the officials or organizations that made them.
The performance figures came from Google, and automatic cleanup may be inappropriate when exact wording is required. That uncertainty remains part of the report because later court orders, technical examinations, audited data, investigative records or field access could change the account.
The checked reporting on Google Brings Meaning-Aware Transcription to More Products identifies two concrete follow-up points: independent accuracy testing across languages and accents and controls that distinguish verbatim mode from rewritten mode. Neither had produced a later attributable outcome before this edition's research cutoff.
The source record also distinguishes the confirmed event—Google calls the model Gemini 3.5 Transcribe.—from the reported consequence—The model supports eighty-five languages and up to three speakers in prerecorded audio.—and from the unresolved boundary: The performance figures came from Google, and automatic cleanup may be inappropriate when exact wording is required. This avoids assigning motive, cause or certainty beyond the cited reporting.
Ars Technica supplied the dated reporting and direct statements used for Google Brings Meaning-Aware Transcription to More Products. Where a government, company, military, hospital or advocacy participant described its own conduct, this article treats that description as an attributed claim rather than independent verification.