Microsoft AI models now available on AI Gateway
Vercel's AI Gateway now supports Microsoft AI models including MAI-Voice-2.1, MAI-Voice-2.1-Flash for speech generation, and MAI-Transcribe-2 Streaming for live transcription. These models are billed at listed rates with no platform markup, offering users control over data and transparency with Microsoft's zero data retention policy.
Vercel and Microsoft AI (MAI) have partnered to make MAI models available on AI Gateway. AI Gateway is one of a limited number of platforms offering access to MAI models.
MAI's focus on security and zero data retention (ZDR) align with Vercel's goal of giving AI Gateway users control over their data and transparency in how it is used for training.
are now supported on AI Gateway. MAI-Voice-2.1 and MAI-Voice-2.1-Flash generate speech, while MAI-Transcribe-2 Streaming returns transcript updates as audio arrives.MAI's latest audio models
AI Gateway bills these models at their listed rates, with no platform fee or markup on inference.
Use in AI SDK 7 with MAI-Voice-2.1-Flash to create a spoken response. The voice name selects Microsoft's Harper voice with the Flash model:generateSpeech
For longer audio such as narration, use .microsoft/mai-voice-2.1
takes a stream of audio chunks and returns transcript updates as the audio arrives. This example assumes is a of 16 kHz, 16-bit PCM audio:streamTranscribemicrophoneStreamReadableStream
Partial transcripts can change as more audio arrives, so replace the displayed text when a new partial arrives.
Explore for speech generation and for live transcription. See the for the full family, or follow the to get started.MAI-Voice-2.1-FlashMAI-Transcribe-2 StreamingMAI model pagespeech quickstart
AI Gateway provides one API for calling models, tracking usage and cost, and viewing request traces. It also supports routing, retries, and failover across available providers.
MAI models on AI Gateway
Get started
Additional resources
() generates expressive speech in 23 languages and keeps a consistent speaker across longer passages. Use it for narration, audiobooks, podcasts, and lessons where delivery matters from start to finish.MAI-Voice-2.1
microsoft/mai-voice-2.1() brings the same multilingual speech generation to lower-latency interactions. It is suited to voice agents, assistants, and spoken replies that need to arrive quickly.MAI-Voice-2.1-Flash
microsoft/mai-voice-2.1-flash() returns partial transcripts while audio is still arriving. Applications can show live captions or follow a conversation as it happens, without waiting for the full recording.MAI-Transcribe-2 Streaming
microsoft/mai-transcribe-2-streaming
Transcribe live audio
Source: original entry ↗
More from Vercel
Follow Vercel to get its new changes in your feed and email digest.
Muse Wonder now available on AI Gateway
Meta's Muse Wonder model is now accessible through Vercel's AI Gateway, supporting coding tasks, multi-step agent workflows, and computer use with a 1M-token context window. Users can access the model via a single API key without creating a separate Meta account and use it for text, image, PDF processing, tool calling, and structured outputs.
Agents can now buy domains with the Vercel CLI
Vercel CLI now enables agents to discover, check availability, and initiate domain purchases through structured commands, with purchase confirmation remaining under user control via interactive terminal or dashboard approval.
New Pro teams default to 30-day deployment retention
Pro teams now default to a 30-day deployment retention period to balance keeping recent history available while reducing storage costs. You can configure a longer retention period in team settings if needed.