Voice and music · by Microsoft AI

MAI-Voice

Narrate long-form text expressively in 23 languages, with emotion styles.

text → speech · speech → profile
Way in
Official API in betaSign up yourself, api key
To start
$Azure free account credit
Runs
In the cloudAlso through Azure, OpenRouter
Checked
Nov 3, 2026Prices as of that day

What it can do 3

Synthesize expressive long-form speech (MAI-Voice-2.1)23 languages; ~550 ms inferencetext → speech$22/1M chars
Synthesize low-latency speech for agents (MAI-Voice-2.1-Flash)~45 ms inference (vendor)text → speech$15/1M chars
Clone a consented voice from a 5-60 s clip (personal voice)Limited Access approval plus an uploaded consent recordingspeech → profilesee Azure pricing

Does the same job

Prices and limits as of Nov 3, 2026, from Microsoft AI’s own pages. They change; check before you rely on them.Not affiliated with Microsoft AI. Not advice.Something wrong?
Made by Jasper Burns

A short note when I ship something new, about once a month. One list for all my projects.