Microsoft AI: MAI-Voice-2.1
Text to speechMAI-Voice-2.1 is Microsoft AI's highest-fidelity, most expressive text-to-speech model. It produces natural, studio-grade speech across 23 languages, with detailed prosody, nuanced expressiveness, and speaker consistency over long-form content.
- Price
- 7,929 Toman per 1,000 characters
- Endpoint
- POST /v1/audio/speech
- Published
- Oct 1, 2026
- Offers
- 97 voices
- Offers
- MP3 · WAV
- Maker
- Microsoft AI
Free plan prices, in Toman at today's rate. Plans and prices
Price on each plan
Per character of the text read, in Toman at today's rate; a text of 1,000 characters is about a minute and a half of speech.
| Plan | Price |
|---|---|
| Free | 7,929 Toman per 1,000 characters |
| Pro | 7,604 Toman per 1,000 characters |
| Max | 7,279 Toman per 1,000 characters |
Voices
97 voices in all: any of them may be asked for by name in the API.
Hear all voicesCall it
curl https://ai.fluxcdn.cloud/api/v1/audio/speech -H "Authorization: Bearer $FLUXCDN_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "microsoft/mai-voice-2.1", "input": "سلام! به FluxCDN خوش آمدید.", "voice": "cs-CZ-Grant:MAI-Voice-2.1", "response_format": "mp3"}' \
--output speech.mp3
