MAI-Voice-2.1-Flash

Not yet independently verified. Every detail here comes from Microsoft AI’s own announcement of 1 October 2026. The accuracy ranking and speed comparisons are Microsoft’s statements; we have not read the ranking site’s own page, and no independent press report had been found at 2026-10-04. We will update this when it can be confirmed, and remove this note.

MAI-Voice-2.1-Flash is the high-volume, low-latency variant of MAI-Voice-2.1, released by Microsoft AI on 1 October 2026. It supports the same 23 languages and cross-language voices as the full model, including voice cloning from a few seconds of reference audio. Microsoft says it can generate 45 seconds of audio with an end-to-end latency of 150 milliseconds, that its inference is 55% faster, and that it is about 60% cheaper than comparable models; those comparisons are Microsoft’s own. It is priced at $15 per million characters. Microsoft positions it as the speaking half of a voice agent, paired with MAI-Transcribe-2-Streaming for listening, for uses such as customer-service lines. It is available through Microsoft Foundry, the MAI Playground, Vercel, Azure Voice Live and OpenRouter. The weights are not published.