MAI-Voice-2.1 pricing vs Sume TTS: what Sume lists, what it does not
MAI-Voice-2.1 costs $22 per 1M characters per Microsoft AI. Sume does not list MAI; its TTS runs Sonic voices at $47.50 per 1M characters.
Sume does not list Microsoft's MAI-Voice-2.1 or MAI-Transcribe-2-Streaming; its text-to-speech route runs Sonic voices and its speech-to-text route is a separate fixed-flag job. Microsoft AI prices MAI-Voice-2.1 at $22 per 1M characters, against Sume TTS at $47.50 per 1M characters ($0.0475 per 1,000), so if you want MAI today you buy it where Microsoft sells it.
Microsoft facts are from its page Our first streaming transcription model debuts at no. 1 on Artificial Analysis (read 2026-10-10). Sume facts are from the Sume OpenAPI contract and the TTS router catalog in the repository.
What Microsoft says it launched
The page names MAI-Transcribe-2-Streaming, MAI-Voice-2.1 and MAI-Voice-2.1-Flash. It says MAI-Voice-2.1 supports 23 languages and 26 locales, and that Flash generates 45 seconds of audio with 150 ms end-to-end latency. It lists availability through Microsoft Foundry, OpenRouter, Vercel and the MAI Playground, with LiveKit coming soon. The transcription model is priced at $0.54 per hour as an introductory rate through year-end.
| Item | Price | Unit |
|---|---|---|
| MAI-Voice-2.1 | $22 | per 1M characters |
| MAI-Voice-2.1-Flash | $15 | per 1M characters |
| MAI-Transcribe-2-Streaming | $0.54 (introductory) | per hour |
| Sume TTS 1.0 | $47.50 ($0.0475 per 1,000) | per 1M characters |
| Sume STT 1.0 | $0.60 ($0.01 per minute) | per hour |
What Sume lists instead
The TTS router accepts sonic-3.6, sonic-3.5, sonic-3, latest (an alias for 3.6) and preview, which is a beta that returns voice_model_mismatch for professional clones. Sume STT takes a public audio URL of up to 600 seconds, returns word timings, and is billed at $0.01 per audio minute. It does not do streaming; a job is a request and a result.
That last point is the real difference with MAI-Transcribe-2-Streaming. A streaming model returns partial text while the speaker is still talking. Sume's job envelope has a sync wait of up to 30 seconds, but that is waiting for one finished result, not a live stream.
How to decide
The price gap is real, but price per character is only part of a voice project. Compare the languages you need, whether your voice has to be a specific person, and whether you need live partials.
- Live captions on a call: a streaming model is the right tool; Sume STT is not.
- Voice-over for finished videos: Sume's TTS plus timeline audio keeps it in one project, and it is $0.0475 per 1,000 characters.
- A specific cloned voice: cloning is app-only on Sume; the API takes a voice id you already have.
- Lowest cost per character: on published prices, the MAI figures are lower.
A caution on the Microsoft numbers
The $0.54 transcription price is described as introductory through year-end, so it may change in January. The latency and accuracy claims are Microsoft's own statements on its own page; I did not measure them and the comparison above does not use them.
Sources
Related posts
More in Comparisons
- Meta skips AI labels: no significant edit, no real-looking human
Meta applies no AI label when its own tools make no significant edit and no photorealistic human. See how Sume trim, crop and dim relate, from Meta's text.
- Meta image expansion or a native 9:16 Sume image for Reels ads?
Meta can expand a creative to 9:16 with generative AI for Reels ads. A native 9:16 image from Sume keeps the composition you chose. When each fits.
- Nano Banana 2.1 or Pro for ad creative? 2.1 at 2K beats Pro at 1K
On Sume, Nano Banana 2.1 at 2K is $0.15, less than Pro at 1K ($0.1875). 2.1 also lists 1:4, 4:1, 1:8 and 8:1 banners that Pro lacks.
- A 10-second Grok Imagine 1.5 Lite clip vs Sume 720p
xAI lists Grok Imagine Video 1.5 Lite at $0.020 a second: $0.20 for one 10-second clip. Sume's grok-imagine-video-1.5 bills $0.125 at 720p.
Written by Sume