Microsoft previews MAI image and voice model variants

Microsoft AI introduced MAI-Image-2.5-Pro and MAI-Voice-2-Flash, expanding its in-house model lineup with options aimed at different quality, speed, and cost requirements. The company said its MAI models are trained on clean, traceable, enterprise-grade data without distillation from third-party models, and are already powering experiences across Bing, PowerPoint, OneDrive, Dynamics 365, and Azure.

MAI-Image-2.5-Pro is now in public preview for high-fidelity image use cases such as hero imagery, detailed editing, and precise in-image text rendering. It is priced at $5 per 1M text input tokens, $8 per 1M image input tokens, and $106 per 1M image output tokens. MAI-Voice-2-Flash is also in public preview, built for high-volume voice experiences where responsiveness matters. Microsoft says Flash is 2x faster than MAI-Voice-2 and 32% cheaper, priced at $15 per 1M characters.

Microsoft said Bing Image Creator is now 100% in-house by default with MAI-Image-2.5, while PowerPoint image-to-image capabilities have reduced GPU costs up to 84% compared with GPT-Image-2. In OneDrive, MAI-Image-2.5 increased save rates by 26%, reduced P95 latency by approximately 25%, and delivered 2.5x greater efficiency under medium-utilization production workloads. MAI-Voice-2-Flash now powers Dynamics 365 Contact Center, reducing GPU costs up to 89%.