Cheap transcription at about half a cent a minute with speaker labels, which makes it an easy pick for meeting and media app developers.

Kenning's take

What it does

Gemini 3.5 Transcribe turns recorded speech into text, detecting more than 85 languages and handling speakers who switch language mid-sentence. It labels speakers, gives word-level timestamps and can be biased towards specialist terms. A live version streams transcription over WebSockets.

• 85+ languages with automatic detection
• Speaker diarisation
• Word-level timestamps
• Custom vocabulary biasing
• Live streaming version
• Up to 1 hour of audio per request

Who it suits. Developers adding transcription to meeting, media or voice apps.

Key facts

Category

AI models

Part of

Google Gemini

Made by

Google

Based in

United States

Launched

2026

How to get it

Gemini API and AI Studio (gemini-3.5-transcribe and gemini-3.5-transcribe-live)

Works on

API

In the UK

Yes

Free plan

Yes, Gemini API free tier

API price

$2.00 per million audio input tokens (about $0.003/min) and $12.00 per million output tokens (about $0.002/min); Live version $3.50 / $21.00

FAQ

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.