# Microsoft AI — Microsoft Build 2026: MAI keynote transcript | Microsoft AI

- Company: Microsoft AI (microsoft.ai)
- Announced: 2026-06-02
- Category: model-update
- Coverage: not counted
- Announcement: yes
- Group: models
- Source: https://microsoft.ai/news/microsoft-build-2026-mai-keynote-transcript/
- Record: https://forck.live/items/10810-microsoft-build-2026-mai-keynote-transcript-microsoft-ai
- Subject: MAI
- Models affected: MAI-Image-2.5, MAI-Image-2.5 Flash, MAI-Transcribe-1.5, MAI-Voice-2, MAI-Voice-2-Flash, MAI-Thinking-1, MAI-Code-1-Flash
- Context window: 256K context window

Microsoft AI announced a family of seven new MAI models at Build 2026, including image, voice, transcription, thinking, and coding models. MAI-Thinking-1 is a 35B active parameter MoE with a 256K context window, achieving 97% on AIME 25 and 53% on SWE Bench Pro. MAI-Code-1-Flash has 5B parameters and achieves 51% on SWE Bench Pro. The models are available on Foundry, Open Router, Fireworks, and Baseten, and are integrated into Microsoft products like PowerPoint, Copilot, Teams, GitHub, and VS Code.

## Evidence

Verbatim from https://microsoft.ai/news/microsoft-build-2026-mai-keynote-transcript/:

> It’s a 35B active parameter MoE with a 256K context window. That means it competes in the medium-sized weight class, where it’s certainly punching above its weight.

---

Record: https://forck.live/items/10810-microsoft-build-2026-mai-keynote-transcript-microsoft-ai
Catalogue: https://forck.live/llms.txt
Current issue: https://forck.live/feed.md