AIME 2026 (AIME26)
A 2026 American Invitational Mathematics Examination snapshot used in frontier-model comparison tables for mathematical reasoning.
Top models on AIME26 — November 21, 2026
As of November 21, 2026, GLM-5.2 leads the AIME26 leaderboard with 99.2% , followed by Beam 1.0 501B-A23B FP8 (98.0%) and Beam 1.0 501B-A23B (97.7%).
GLM-5.2
Z.AI
Reported by Z.AIBeam 1.0 501B-A23B FP8
Reflection AI
Reported by Reflection AIBeam 1.0 501B-A23B
Reflection AI
Reported by Reflection AI34 modelsMathematics6% of Math reference weightCurrentUpdated November 21, 2026
Leaderboard (34 models)
None of the 34 results on this page has a documented run by the benchmark or an independent evaluator: 31 reported by the labs that built the models and 3 reported in another publisher’s document.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | GLM-5.2Z.AIReported by Z.AI | 99.2% | Reported by Z.AI | Not reported | Open |
| 2 | Beam 1.0 501B-A23B FP8Reflection AIReported by Reflection AI | 98.0% | Reported by Reflection AI | Not reported | Open |
| 3 | Beam 1.0 501B-A23BReflection AIReported by Reflection AI | 97.7% | Reported by Reflection AI | Not reported | Open |
| 4 | Beam 1.0 501B-A23B NVFP4Reflection AIReported by Reflection AI | 97.5% | Reported by Reflection AI | Not reported | Open |
| 5 | InklingThinking Machines LabReported by Thinking Machines Lab | 97.1% | Reported by Thinking Machines Lab | Not reported | Open |
| 6 | A.X K2SK TelecomReported by SK Telecom | 97.1% | Reported by SK Telecom | Not reported | Open |
| 7 | Kimi K2.6Moonshot AIReported by Moonshot AI | 96.4% | Reported by Moonshot AI | Not reported | Open |
| 8 | Ternary Bonsai 2 27BPrism MLReported by Prism ML | 95.8% | Reported by Prism ML | Not reported | Open |
| 9 | GLM-5Z.AIAnother lab’s comparison table | 95.8% | Another lab’s comparison table | Not reported | Open |
| 10 | Kimi K2.5Moonshot AIReported by Moonshot AI | 95.8% | Reported by Moonshot AI | Not reported | Open |
| 11 | Solar Open 2UpstageReported by Upstage | 95.7% | Reported by Upstage | Not reported | Open |
| 12 | Inkling-SmallThinking Machines LabReported by Thinking Machines Lab | 95.5% | Reported by Thinking Machines Lab | Not reported | Open |
| 13 | Qwen3.6 PlusAlibabaReported by Alibaba | 95.3% | Reported by Alibaba | Not reported | Closed |
| 14 | GLM-5.1Z.AIReported by Z.AI | 95.3% | Reported by Z.AI | Not reported | Open |
| 15 | Solar Pro 4UpstageReported by Upstage | 95.3% | Reported by Upstage | Not reported | Closed |
| 16 | Claude Opus 4.5AnthropicAnother lab’s comparison table | 95.1% | Another lab’s comparison table | Not reported | Closed |
| 17 | Ripple-65B PreviewRicursive IntelligenceReported by Ricursive Intelligence | 95.0% | Reported by Ricursive Intelligence | Not reported | Open |
| 18 | Muse Glimmer 30BMetaReported by Meta | 94.7% | Reported by Meta | Not reported | Open |
| 19 | MAI-Thinking-1MicrosoftReported by Microsoft | 94.5% | Reported by Microsoft | Not reported | Closed |
| 20 | Qwen3.6-27BAlibabaReported by Alibaba | 94.1% | Reported by Alibaba | Not reported | Open |
| 21 | Qwen3.5 397BAlibabaReported by Alibaba | 93.3% | Reported by Alibaba | Not reported | Open |
| 22 | Ling 3.0 FlashInclusionAIReported by InclusionAI | 93.2% | Reported by InclusionAI | Not reported | Open |
| 23 | Qwen3.6-35B-A3BAlibabaReported by Alibaba | 92.7% | Reported by Alibaba | Not reported | Open |
| 24 | K-EXAONE 2.0LG AI ResearchReported by LG AI Research | 92.3% | Reported by LG AI Research | Not reported | Open |
| 25 | Spark-X2.6-4BSparkLLMReported by SparkLLM | 91.2% | Reported by SparkLLM | Not reported | Open |
| 26 | MiniMax M2.7MiniMaxAnother lab’s comparison table | 89.8% | Another lab’s comparison table | Not reported | Open |
| 27 | ZAYA1-8BZyphraReported by Zyphra | 89.1% | Reported by Zyphra | Not reported | Open |
| 28 | MiniCPM5-2BOpenBMBReported by OpenBMB | 86.5% | Reported by OpenBMB | Not reported | Open |
| 29 | Gemma 4 12BGoogleReported by Google | 77.5% | Reported by Google | Not reported | Open |
| 30 | ZAYA1-74B-PreviewZyphraReported by Zyphra | 76.4% | Reported by Zyphra | Not reported | Open |
| 31 | LongCat-Flash-Lite-SparseMeituanReported by Meituan | 65.7% | Reported by Meituan | Not reported | Open |
| 32 | LFM2.5-8B-A1BLiquidAIReported by LiquidAI | 50.0% | Reported by LiquidAI | Not reported | Open |
| 33 | MiniCPM5-1BOpenBMBReported by OpenBMB | 40.4% | Reported by OpenBMB | Not reported | Open |
| 34 | LLaDA2.2-miniInclusionAIReported by InclusionAI | 35.0% | Reported by InclusionAI | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
According to BenchLM.ai, GLM-5.2 leads the AIME26 benchmark with a score of 99.2%, followed by Beam 1.0 501B-A23B FP8 (98.0%) and Beam 1.0 501B-A23B (97.7%). The top models are clustered within 1.5 points, suggesting this benchmark is nearing saturation for frontier models.
34 models have been evaluated on AIME26. The benchmark falls in the Mathematics category. BenchAlign v6.0 gives AIME26 6% of the Math reference weight, so it moves the Math leaderboard and the overall ranking. Reference weights are relative weights in the calibrated model, not fixed shares of a score.
About AIME26
Year
2026
Tasks
Competition math problems
Format
Short-answer mathematics
Difficulty
Olympiad-style mathematics
AIME-style benchmarks remain one of the fastest ways to separate top reasoning models on olympiad-style math. AIME 2026 is a newer contest-year snapshot than the legacy AIME rows already tracked on BenchLM.
Freshness and provenance
Version
AIME26 2026
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does AIME26 measure?
A 2026 American Invitational Mathematics Examination snapshot used in frontier-model comparison tables for mathematical reasoning.
Which model scores highest on AIME26?
GLM-5.2 by Z.AI currently leads with a score of 99.2% on AIME26.
How many models are evaluated on AIME26?
34 AI models have published results on AIME26 in the BenchLM catalog.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.