Microsoft MAI-Thinking-1
B Tier · 7.5/10
Microsoft's first in-house reasoning model -- launched 2026-06-02 at Build as the flagship of seven new MAI models. 35B-active / ~1T-total sparse Mixture-of-Experts, 256K context. AIME 2025 97.0%, matches leading models on SWE-Bench Pro, and beat Claude Sonnet 4.6 in human-preference testing. Available on Microsoft Foundry + OpenRouter / Fireworks / Baseten
Score Breakdown
Benchmark Scores
Benchmarks for MAI-Thinking-1 (vendor-published 2026-06-02; third-party verification pending)
| Benchmark | Description | Score | |
|---|---|---|---|
| AIME 2025 | 97% | ||
| AIME 2026 | 94.5% |
Last updated: 2026-06-02
The Good and the Bad
What we like
- +Microsoft's first in-house frontier-class reasoning model -- removes the OpenAI dependency for the reasoning tier the same way MAI-Voice/Image/Transcribe did for speech and vision
- +Strong published reasoning numbers: AIME 2025 97.0% and AIME 2026 94.5%, plus it 'matches leading models on key software engineering benchmarks' (SWE-Bench Pro) at a medium model size
- +Sparse MoE design (35B active of ~1T total) is built for cost-efficiency -- Microsoft is positioning it as strong reasoning per dollar rather than a max-size flagship
- +256K context + Chat Completions API compatibility makes it a near drop-in for existing reasoning workloads, and it shipped with day-one availability on OpenRouter, Fireworks, and Baseten
What could be better
- −Pricing not disclosed at launch -- you cannot yet model cost vs Claude / Gemini / GPT reasoning tiers without going through a third-party inference provider's published rate
- −Foundry access is private preview at launch (MAI Playground public preview is 'coming soon') -- broad self-serve evaluation is not open yet
- −Microsoft's benchmark and preference numbers are self-reported; independent AIME / SWE-Bench Pro confirmation typically lags announcement by several weeks
- −No open weights -- this is an API/Foundry model, not a self-hostable open-weight release like Phi or the local-LLM tier
Pricing
Microsoft Foundry
- ✓Launched 2026-06-02 (Microsoft Build) -- private preview on Foundry at launch
- ✓256K-token context window
- ✓Chat Completions API compatible
- ✓Public preview on MAI Playground coming soon
Third-party inference (OpenRouter / Fireworks / Baseten)
- ✓Generally available at launch through OpenRouter, Fireworks, and Baseten
- ✓Pay-as-you-go at each provider's per-token rate
- ✓No first-party consumer subscription -- API/developer access only
Known Issues
- LAUNCH (2026-06-02, Microsoft Build): MAI-Thinking-1 is the flagship of the 'seven new MAI models' wave (alongside MAI-Code-1-Flash, MAI-Image-2.5 + Flash, MAI-Transcribe-1.5, MAI-Voice-2 + Flash). Microsoft describes it as 'a medium-sized model that stands among the strongest models in its weight class.' Vendor-published: AIME 2025 97.0%, AIME 2026 94.5%, matches leading models on SWE-Bench Pro, and in human-preference testing 'users preferred MAI-Thinking-1 over Claude Sonnet 4.6.' Architecture: 35B-active, ~1T-total sparse Mixture-of-Experts; 256K context; Chat Completions API compatible. Availability: private preview on Microsoft Foundry at launch + GA through OpenRouter / Fireworks / Baseten; MAI Playground public preview coming soon.Source: Microsoft AI (microsoft.ai/news/introducing-mai-thinking-1/, microsoft.ai/news/building-a-hillclimbing-machine-launching-seven-new-mai-models/) · 2026-06-02
- Self-reported benchmarks pending third-party verification. Microsoft's AIME and SWE-Bench Pro figures and the Sonnet 4.6 preference-win claim are first-party. Treat as vendor numbers until Artificial Analysis / LMArena-style independent confirmation lands (typically 4-8 weeks).Source: Microsoft model announcement · 2026-06
Best for
Azure / Microsoft Foundry shops that want a first-party reasoning model without an OpenAI dependency, and developers who want a cost-efficient reasoning tier (sparse MoE, 256K context) accessible today through OpenRouter, Fireworks, or Baseten.
Not for
Consumer users who want a chat UI (there is no claude.ai-style website -- this is API/Foundry only), teams that need open weights for self-hosting (use the local-LLM tier), or anyone who needs published per-token pricing before committing.
Our Verdict
MAI-Thinking-1 is the most strategically significant entry in Microsoft's June 2, 2026 Build model wave: it is Microsoft's first in-house reasoning model, and the published numbers (AIME 2025 97.0%, SWE-Bench Pro parity with leading models, a human-preference win over Claude Sonnet 4.6) are genuinely competitive for a 35B-active model. Combined with MAI-Code, MAI-Image, MAI-Voice, and MAI-Transcribe, it completes Microsoft's in-house coverage of every major modality and removes the last big OpenAI dependency at the reasoning tier. The open questions are pricing (undisclosed) and independent verification of the benchmarks -- but for Foundry customers this is now a real first-party reasoning option, and the day-one OpenRouter/Fireworks/Baseten availability makes it easy to try.
Sources
- Microsoft AI: Introducing MAI-Thinking-1 (2026-06-02) (accessed 2026-06-02)
- Microsoft AI: Building a hill-climbing machine -- launching seven new MAI models (2026-06-02) (accessed 2026-06-02)
Explore more Microsoft MAI-Thinking-1 rankings
Deeper leaderboards, benchmarks, task-specific tier lists, and status/pricing pages for Microsoft MAI-Thinking-1.
The Tier List Tuesday
Weekly newsletter: tier movers, new entrants, and the VS of the week. Built from our daily AI-tool sweeps. No spam, unsubscribe anytime.
Alternatives to Microsoft MAI-Thinking-1
Claude (Anthropic)
Anthropic's flagship LLM family. **Claude Opus 5 launched 2026-07-24** and is now the default model on Claude Max and the strongest model on Claude Pro -- same $5/$25 per 1M as Opus 4.8, but Anthropic says it lands within 0.5% of Fable 5 on CursorBench at half the cost. **Sonnet 5's $2/$10 per 1M is now permanent** -- Anthropic cancelled the 2026-09-01 rise to $3/$15 and made the launch rate standard -- and it stays the default on Free/Pro. **Claude Fable 5.1 and Mythos 5.1 launched 2026-09-01** and now top the range -- same $10/$50 per 1M as Fable 5, with the saving delivered entirely through a 4x cheaper cache read ($1 -> $0.25/MTok), so it is 25-45% cheaper only if your workload reuses cached context. **From 2026-08-14 future Claude models watermark their text output globally** (SynthID-Text; no extra tokens, no price or speed change, no identifying information -- detector API not shipped yet), and the **legacy Workbench plus the experimental prompt-tools APIs retired 2026-08-17**
Claude Mythos 5.1
Anthropic's trusted-access frontier model. **Mythos 5.1 launched 2026-09-01** alongside Fable 5.1, and Anthropic now states outright that they are **the same model with different safeguards** -- the 5.1-cycle gap is 60.9% vs 55.8% on Terminal-Bench 4.0, which Anthropic attributes to safeguard interventions rather than capability and expects to shrink. Originally launched June 9, 2026 alongside Claude Fable 5. Suspended June 12 by a US export-control order, then PARTIALLY RESTORED July 1, 2026 (US government lifted controls June 30): Mythos 5 is back for a set of US organizations with government approval, while Anthropic works to re-expand the broader Glasswing program. Public Fable 5 returned globally the same day. Gated to Project Glasswing orgs + select biology researchers.
Gemini (Google)
Google's LLM with deep Google Workspace integration, 2M token context window, and native code execution -- **Gemini 3.8 Flash launched 2026-09-02** -- the third Flash release in six weeks -- at the SAME $0.75/$3.75 per 1M intro rate as 3.7 Flash and, critically, **on the same 2026-12-31 expiry** (then $1.50/$7.50), so the discount window did not reset. HLE-Verified 54.9%; Google warns it uses more tokens per task because it 'works harder', so cost per task can rise at an unchanged rate. **Gemini 3.8 Flash Cyber** is trusted-defenders-only via the new **Fairwind Program**. Gemini 3.7 Flash (2026-08-13) remains fully supported for efficiency-first work. Gemini 3.5 Pro STILL delayed and partner-testing-only (Bloomberg, 7/16 -- coding shortfalls, no ship date), so Flash keeps shipping while the Pro line stalls. Gemini 4 pre-training underway. **Two new models in Aug 2026: Gemini 3.5 Transcribe (2026-08-26) speech-to-text and Gemini Omni 1.1 Flash (2026-08-27) production video with 40s scene extension, start/end frames and 4K** -- neither shipped with a public rate card
Grok
SpaceXAI's irreverent chatbot with a direct line to X/Twitter -- and now **Grok 4.6 (launched 2026-08-12)**, focused on long-running agents and interactive/visual work, at the same $2/$6 per 1M as Grok 4.5 (fast variant 2x). xAI says it matches GPT-5.6 Sol on the AA Intelligence Index (61), though Sol still leads it on DeepSWE and Terminal-Bench. **Grok Bot** (8/11, early beta) adds always-on agents with their own cloud computer. **Grok 4.6 reached GitHub Copilot on 2026-08-14** across all five paid Copilot SKUs (off by default for orgs), adding to its day-one Cursor availability. Grok 4.3 remains the value tier at $1.25/$2.50
Muse Spark (Meta)
Meta's frontier model line from its Superintelligence Lab -- Muse Spark 1.1 (2026-07-09) adds substantially better coding, 1M-token multi-agent orchestration, and Meta's first paid developer API (Meta Model API, public preview)
GPT-Rosalind (OpenAI)
OpenAI's first domain-specific model -- life sciences, drug discovery, translational medicine. Launched 2026-04-16 as a Trusted Access research preview. Launch partners: Amgen, Moderna, Allen Institute, Thermo Fisher. Paired with a Life Sciences Codex plugin (50+ scientific tool integrations)
GPT-5.4-Cyber (OpenAI)
OpenAI's defensive-cybersecurity variant of GPT-5.4, launched 2026-04-16. Lowered refusal boundary for security-research tasks and native binary reverse-engineering. Access gated via Trusted Access for Cyber (TAC) program -- thousands of verified defenders, hundreds of teams, no public pricing. On **2026-08-17 OpenAI published its first dedicated post on the Hugging Face incident**, conceding it 'underestimated the real-world cyber capabilities of our AI models' and confirming it now releases cyber capabilities only to trusted defenders. **On 2026-09-01 OpenAI confirmed GPT-6 Astra meets the Critical cyber threshold -- the first model it has ever designated at that level -- and on 2026-09-03 committed $1B to Daybreak for Frontline Defenders**
Hunyuan 3 (Tencent Hy3)
Tencent's Hy3 reached GA 2026-07-06 (upgraded from the April preview) -- 295B total / 21B active MoE, 256K context, now Apache 2.0 open weights on HuggingFace + ModelScope with the EU/UK/South Korea restriction lifted. ~90% agent-task completion on Tencent's internal apps; API via Tencent Cloud TokenHub. Integrated into Yuanbao, WeChat, QQ
MiMo (Xiaomi)
Xiaomi's MiMo-V2.5 family launched 2026-04-22 -- Pro (1T total / 42B active MoE, 1M context, native vision+audio reasoning), Multimodal base, TTS (3 sub-models: base, VoiceDesign, VoiceClone), and ASR (open-source, English + Chinese + major dialects). Full voice pipeline for the agent era. Extra-charge 1M-context tier removed at launch