Microsoft MAI models (series)
MAI is the in-house model family of Microsoft AI, the group led by Mustafa Suleyman. The family entry point lists each model in order, with its category, the availability its own announcement states, and the claims Microsoft makes. Benchmark figures are Microsoft’s; no independent evaluation is recorded here.
Lineage
| Date | Release | Notes |
|---|---|---|
| 2026-04-02 | MAI-Transcribe-1, MAI-Voice-1, MAI-Image-2 | First generation, launched in Microsoft Foundry ([[../../../RESOURCES/COMPANIES/microsoft |
| 2026-06 (Build) | MAI-Code-1-Flash | Coding model for GitHub Copilot; the first MAI coding line |
| 2026-09-04 | MAI Image 2.6 | Post “Pushing the quality-cost frontier with MAI Image 2.6” |
| 2026-11-02 | MAI-Image-2.7 | Foundry catalogue version date |
| 2026-11-17 | Ignite batch (below) | Seven models announced; eighth item on the Ignite page is hardware |
The Ignite 2026 batch (2026-11-17)
| Model | Category | Stated availability | Microsoft’s headline claim | Note |
|---|---|---|---|---|
| MAI Code 2 | Coding | GitHub Copilot; Foundry not stated | 98% of GPT-6.1 Sol’s performance at 40% of its relative cost | [[./mai-code-2 |
| MAI Code 2 Flash | Coding (high volume) | GitHub Copilot; limited-access preview in Foundry | Comparable to GPT-6 Luna at 65% of its relative cost | [[./mai-code-2-flash |
| MAI Cyber 2 Flash | Security (agentic defence) | Exclusive to Microsoft security products | Beats GPT-6 Astra on CyberGym-E2E (92.3% vs 91.5%) | [[./mai-cyber-2-flash |
| MAI Image 2.7 | Image generation and editing | Foundry, preview | ”#1 on Artificial Analysis image editing” | [[../image-generation/mai-image-2-7 |
| MAI Realtime 1 | Real-time speech-to-speech | MAI Labs trial; Foundry limited-access preview | Outperforms GPT-Realtime 2.1 | [[../audio/mai-realtime-1 |
| MAI Transcribe 2 | Speech-to-text | Generally available in Foundry | Dragon Copilot Transcribe, built on it, halves multilingual error rates | [[../audio/mai-transcribe-2 |
| Decision 1.1 | Routing and decision | Not confirmed in the primary text | 50 times faster than GPT-6.1 Sol | [[./decision-1-dot-1 |
The Ignite page counts seven models but lists eight items: the eighth, Maia 200, is an AI accelerator chip and is not part of the family.
Source limits
- Every performance and cost figure is Microsoft’s own. Cost claims are relative percentages of competitors’ cost, not published prices.
- Benchmark conditions differ between posts. Each model note records the conditions its own post states.
- The announcement pages were read in summary form; the full text of some posts was not available to the review.
- Secondary outlets give different model-name spellings and availability details. This series note follows the Microsoft pages.
Related
- Microsoft (company note, with the earlier MAI timeline)
- microsoft-azure and microsoft-foundry-note (where most MAI models are hosted)
Sources
- Microsoft Ignite 2026: seven new world-class models (2026-11-17)
- MAI news index (2026-11-22)
- Microsoft Foundry catalogue cards for MAI-Transcribe-2 and MAI-Image-2.7 (fetched 2026-11-22)