Google's Gemini 3.5 Transcribe supports 85 languages with real-time correction
The Decoder · rss · 2026-08-27
Google launched Gemini 3.5 Transcribe, supporting speech-to-text in over 85 languages. It features real-time removal of filler words and correction of slips of the tongue, achieving a 4.0% word error rate in streaming mode with 70% lower latency than its predecessor, Chirp 3. It also supports function calling to hand off tasks to other Gemini models.
More from Models
- Nat Lambert: wave of AI releases rumored — thinky, muse spark, Fable 5.5, OpenAI — natolambert · 2026-10-15
- Musk amplifies dev's verdict: Grok Bot easier to use than Codex and Claude — elonmusk · 2026-10-15
- Rumor: New Anthropic model imminent, apparently behind a batch of high-severity Chrome CVEs — matthew_d_green · 2026-10-15
- StartLux-Decision debuts RL-trained calibrated decision models, 27B scores 72.8 — startlux-models · 2026-10-15
- Meta's Muse Code slammed for harsh quotas: 21% left after 5 hours — ThomasScialom · 2026-10-15
- Claude Code error mentions 'Fable 5.1' and 'Opus 5', hinting at unreleased models — adonis_singh · 2026-10-15