Capability
52.3/100
field median 58.1
#116 of 232 ranked models
MAI-Thinking-1
Released Jun 2, 2026 — see all recent releases
Decision reading
MAI-Thinking-1 scores 52.3 out of 100 and ranks #116 of 232. This profile shows 14 source-displayable benchmark rows; its strongest eligible category is Instruction Following at #1. No comparable first-party API price is published in the catalog.
Data as of September 4, 2026 · How the score is built
Share or export
Strongest published evidence
Instruction Following ranks #1. A well-rounded choice across a range of tasks.
Validate before choosing
14 published rows leave some tracked benchmark slots empty. No comparable first-party API token rate is published.
Decision snapshot
Each value carries a field reference instead of floating alone. Markers compare this model with the current ranked and priced catalog; they are not absolute quality thresholds.
Price
Not listed
input median $1
No comparable first-party hosted token rate
Speed
Not measured
field median 92 tok/s
Time to first token not measured
Context
256Ktokens
field median 256,000
Reported for this model; direct source link not stored
Capability shape
Each axis shows percentile within that category’s eligible cohort. The comparison outline is the median of the six nearest public-score peers; a collapsed vertex means the category is not rank-eligible.
MAI-Thinking-1 category percentile values
- Agentic65th percentile
- Coding68th percentile
- ReasoningNot eligible
- Knowledge63rd percentile
- MathNot eligible
- MultilingualNot eligible
- MultimodalNot eligible
- Instruction following100th percentile
The dashed outline is median of 6 nearest peers.
Eligible category ranks
- Agentic#53/151
- Coding#60/183
- ReasoningNot ranked
- Knowledge#68/181
- MathNot ranked
- MultilingualNot ranked
- MultimodalNot ranked
- Inst. Following#1/120
Top decileTop quartileMid-fieldNot eligible
How much of this is verified
Coverage is split by category so a strong number never hides a thin evidence base. Verified means the row is tied to a published source; provisional rows remain visible but separate.
- Agentic1/1 verified
- Coding4/4 verified
- Reasoning1/1 verified
- Knowledge4/4 verified
- Math3/3 verified
- MultilingualNot measured
- MultimodalNot measured
- Inst. Following1/1 verified
Verified sourceProvisionalNot measured
Spec sheet
Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.
- API model ID
- Not published
- Context window
- 256K
- Maximum output
- Not sourced yet
- Knowledge cutoff
- Not sourced yet
- Input modalities
- Not sourced yet
- Output modalities
- Not sourced yet
- Parameters
- Not disclosed by the provider
- Availability
- Not sourced yet
- Cloud regions
- Not tracked yet
- Lifecycle
- Current
- API capabilities
- Tool calling, structured outputs, and batch support are not tracked yet
- Prompt caching
- Not documented in the pricing record
- Self-host
- Weights are not published
- Rate limits
- Not tracked yet
Category score record
Scores and ranks appear only where published evidence can be displayed. The table keeps the score, weight, cohort, and evidence state together.
| Category | Score | Rank | Percentile | Weight | Benchmarks | Evidence |
|---|---|---|---|---|---|---|
| AgenticRank #53 of 151Percentile 65thWeight 22%1 benchmarkVerified | 51.7 | #53 of 151 | 65th | 22% | 1 benchmark | Verified |
| CodingRank #60 of 183Percentile 68thWeight 20%4 benchmarksVerified | 51.8 | #60 of 183 | 68th | 20% | 4 benchmarks | Verified |
| ReasoningWeight 17%1 benchmarkVerified | Score pending | Not ranked | Not available | 17% | 1 benchmark | Verified |
| KnowledgeRank #68 of 181Percentile 63rdWeight 12%4 benchmarksVerified | 53.3 | #68 of 181 | 63rd | 12% | 4 benchmarks | Verified |
| MathRank Not rankedWeight 5%3 benchmarksVerified | 73.4 | Not ranked | Not available | 5% | 3 benchmarks | Verified |
| MultilingualWeight 7%0 benchmarksNot measured | Not measured | Not ranked | Not available | 7% | 0 benchmarks | Not measured |
| MultimodalWeight 12%0 benchmarksNot measured | Not measured | Not ranked | Not available | 12% | 0 benchmarks | Not measured |
| Inst. FollowingRank #1 of 120Percentile 100thWeight 5%1 benchmarkVerified | 94.7 | #1 of 120 | 100th | 5% | 1 benchmark | Verified |
Benchmark ledger
Coding opens by default. The marker compares each value with the best source-verified result in the catalog; provisional leaders do not set the reference. Expand the remaining categories for every published row.
Coding4 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| SWE-bench Pro | Score52.8% | Versus best verified row Best verified: Claude Fable 5.1 · 81.2% | Gap28.4 behind | WeightWeighted 25% | Provider exact |
| SWE-bench VerifiedSoftware Engineering Benchmark Verified | Score73.5% | Versus best verified row Best verified: Claude Opus 5 · 96% | Gap22.5 behind | WeightWeighted 10% | Provider exact |
| LiveCodeBench v6 | Score87.7% | Versus best verified row Best verified: Sakana Fugu-Ultra · 93.2% | Gap5.5 behind | WeightDisplay only | Provider exact |
| Terminal-Bench 2.0 | Score46.0% | Versus best verified row Best verified: GPT-5.6 Sol · 91.9% | Gap45.9 behind | WeightDisplay only | Provider exact |
Agentic1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Terminal-Bench 2.0 | Score46% | Versus best verified row Best verified: GPT-5.6 Sol · 91.9% | Gap45.9 behind | WeightWeighted 30% | Provider exact |
Reasoning1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| Graphwalks BFS 128KGraphwalks BFS 0K-128K | Score90% | Versus best verified row Best verified: MAI-Thinking-1 · 90% | GapBest verified | WeightDisplay only | Provider exact |
Knowledge4 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| MMLU-ProMassive Multitask Language Understanding Professional | Score85% | Versus best verified row Best verified: Qwen3.7 Max · 89.6% | Gap4.6 behind | WeightWeighted 20% | Provider exact |
| GPQAGraduate-Level Google-Proof Q&A | Score84.2% | Versus best verified row Best verified: GPT-6 Astra · 96% | Gap11.8 behind | WeightWeighted 7% | Provider exact |
| SimpleQAMeasuring Short-Form Factuality in Large Language Models | Score31% | Versus best verified row Best verified: DeepSeek V4 Pro 0813 · 57.9% | Gap26.9 behind | WeightWeighted 5% | Provider exact |
| GPQA-DGPQA Diamond | Score84.2% | Versus best verified row Best verified: GPT-6 Astra · 96.0% | Gap11.8 behind | WeightDisplay only | Provider exact |
Math3 rows
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| AIME26AIME 2026 | Score94.5% | Versus best verified row Best verified: GLM-5.2 · 99.2% | Gap4.7 behind | WeightWeighted 25% | Provider exact |
| HMMT Feb 2026Harvard-MIT Mathematics Tournament February 2026 | Score84.9% | Versus best verified row Best verified: Qwen3.7 Max · 97.1% | Gap12.2 behind | WeightWeighted 25% | Provider exact |
| AIME 2025American Invitational Mathematics Examination 2025 | Score97% | Versus best verified row Best verified: MAI-Thinking-1 · 97% | GapBest verified | WeightDisplay only | Provider exact |
Inst. Following1 row
| Benchmark | Score | Versus best verified row | Gap | Weight | Evidence |
|---|---|---|---|---|---|
| IFBenchInstruction Following Benchmark | Score85% | Versus best verified row Best verified: MAI-Thinking-1 · 85% | GapBest verified | WeightWeighted 70% | Provider exact |
Lineage
The sequence follows explicit supersedes links. A successor's displayed score stays at least 0.1 points above its predecessor; raw benchmark rows do not move. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.
Jun 2, 2026 · you are here
MAI-Thinking-1Score 52.3 · Price not listed
1
How to read this profile
The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.
MAI-Thinking-1 ranks #116 of 232 on the public leaderboard with a score of 52.27/100. It does not yet have enough sourced coverage for a verified position.
MAI-Thinking-1 is a proprietary model with a 256K context window. It uses an explicit reasoning mode, which can improve complex problem solving while adding latency and token use.
Official exact-value snapshot from Microsoft's June 2026 MAI-Thinking-1 technical report and model page. MAI describes the model as a 35B-active, roughly 1T-total-parameter MoE trained from scratch; BenchLM maps the public STEM, coding, instruction-following, and long-context values that match existing schema keys.
14 of 422 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.
Its strongest eligible category is Instruction Following at #1, while its lowest eligible position is Knowledge at #68. a well-rounded choice across a range of tasks.
Radar
Full release historyMAI-Thinking-1 release history
MAI-Thinking-1 released
Microsoft · Model release
Source confirmed
Radar confirmed these at the source. Already tracking MAI-Thinking-1? See the free Radar Brief for what changes next.
Frequently asked questions
How does MAI-Thinking-1 perform overall in AI benchmarks?
MAI-Thinking-1 ranks #116 out of 232 models on the public BenchAlign leaderboard, with a score of 52.27/100. Its evidence status is Estimated, and this profile shows 14 source-displayable benchmark rows. The label describes evidence depth, not a provider quality claim; inspect category rows before choosing a workload.
Is MAI-Thinking-1 good for knowledge and understanding?
MAI-Thinking-1 ranks #68 out of 181 eligible models for knowledge and understanding, with a public category score of 53.3/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is MAI-Thinking-1 good for coding and programming?
MAI-Thinking-1 ranks #60 out of 183 eligible models for coding and programming, with a public category score of 51.8/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is MAI-Thinking-1 good for mathematics?
MAI-Thinking-1 has source-displayable benchmark coverage for mathematics, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is MAI-Thinking-1 good for reasoning and logic?
MAI-Thinking-1 has source-displayable benchmark coverage for reasoning and logic, but the public category table does not assign it a rank there. The individual rows remain available for inspection. A missing category position means the evidence threshold was not met; it does not convert the model's unmeasured work into a zero.
Is MAI-Thinking-1 good for agentic tool use and computer tasks?
MAI-Thinking-1 ranks #53 out of 151 eligible models for agentic tool use and computer tasks, with a public category score of 51.7/100. Higher-ranked alternatives are available for workloads where this category decides the choice. Check the underlying rows before treating the aggregate as a workload guarantee.
Is MAI-Thinking-1 good for instruction following?
MAI-Thinking-1 ranks #1 out of 120 eligible models for instruction following, with a public category score of 94.7/100. That places it in the current top ten for this category. Check the underlying rows before treating the aggregate as a workload guarantee.
Does MAI-Thinking-1 have full benchmark coverage on BenchLM?
No. MAI-Thinking-1 currently has 14 source-displayable rows across 422 tracked benchmark slots. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.
What is the context window size of MAI-Thinking-1?
MAI-Thinking-1 has a reported context window of 256K in the exact-model catalog record. The value stays visible, but the profile marks its source link as unavailable instead of presenting it as directly documented. Maximum output length remains separate because providers often publish a different limit.
Compare MAI-Thinking-1 with every tracked model410 comparisons
Last updated September 4, 2026. Runtime fields remain blank until a sourced snapshot exists.
Watch MAI-Thinking-1 in the weekly brief
Get one weekly email when material rank, price, availability, or benchmark evidence changes are worth revisiting.
Read a sample issueJoin 2,000+ readers.