AssemblyAI

Universal-3.5 Pro

AssemblyAI's most accurate batch transcription model: 18 languages, native code switching and speaker diarization.

USD per hour of audio

$0.21

Verified Oct 27, 2026

#6 cheapest of 10 in Speech-to-text

Capabilities, explained

What this API can do, and what each capability means for your project.

Real-time streaming

—

Transcribes live audio with low delay.

Speaker labels

—

Identifies different speakers.

Languages

—

Number of supported languages.

Word timestamps

—

Start/end time for each word, needed for subtitles.

Open weights

—

The model weights are downloadable, so you can self-host it or switch hosting providers without changing models.

Only one price recorded so far. The chart appears once the price changes.