Knowledge · Benchmark profile
A comprehensive multiple-choice question answering test covering 57 tasks including elementary mathematics, US history, computer science, law, and more. Tests knowledge across diverse academic subjects from high school to professional level.
Data verified 27 Jul 2026 · Methodology 1.8.0
Benchmark score on Massive Multitask Language Understanding
o1 leads at 91.8%, followed by GPT-4.1 (90.2%) and DeepSeek V4 Pro Base (90.1%).
8 ranked models · higher is better · labels show rank and score
| Model | Rank | Provider | Score |
|---|---|---|---|
| o1 | #1 | OpenAI | 91.8% |
| GPT-4.1 | #2 | OpenAI | 90.2% |
| DeepSeek V4 Pro Base | #3 | DeepSeek | 90.1% |
| DeepSeek V4 Flash Base | #4 | DeepSeek | 88.7% |
| GPT-4.1 mini | #5 | OpenAI | 87.5% |
| Trinity-Large-Preview | #6 | Arcee AI | 87.2% |
| o3-mini | #7 | OpenAI | 86.9% |
| GPT-4.1 nano | #8 | OpenAI | 80.1% |
One best score per model · higher is better
| Rank | Model | Provider | License | Evidence use | Score |
|---|---|---|---|---|---|
| #1 | o1 o1 | OpenAI | closed | Estimated reference | 91.8% |
| #2 | GPT-4.1 gpt-4-1 | OpenAI | closed | Reference only | 90.2% |
| #3 | DeepSeek V4 Pro Base deepseek-v4-pro-base | DeepSeek | open | Estimated reference | 90.1% |
| #4 | DeepSeek V4 Flash Base deepseek-v4-flash-base | DeepSeek | open | Estimated reference | 88.7% |
| #5 | GPT-4.1 mini gpt-4-1-mini | OpenAI | closed | Estimated reference | 87.5% |
| #6 | Trinity-Large-Preview trinity-large-preview | Arcee AI | open | Estimated reference | 87.2% |
| #7 | o3-mini o3-mini | OpenAI | closed | Reference only | 86.9% |
| #8 | GPT-4.1 nano gpt-4-1-nano | OpenAI | closed | Estimated reference | 80.1% |
The top of this snapshot is led by o1 at 91.8%; third place is 1.7 points behind. The top-8 spread is 11.7 points.
About Massive Multitask Language Understanding
A comprehensive multiple-choice question answering test covering 57 tasks including elementary mathematics, US history, computer science, law, and more. Tests knowledge across diverse academic subjects from high school to professional level. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.
Open benchmark source ↗FAQ
A comprehensive multiple-choice question answering test covering 57 tasks including elementary mathematics, US history, computer science, law, and more. Tests knowledge across diverse academic subjects from high school to professional level.
o1 by OpenAI currently leads with 91.8%.
8 models in the LuminaBench cohort have a qualifying score on this benchmark.
No. This benchmark is display-only and does not enter the overall Lumina composite.
Related
2026 · 21 results · Reference
Artificial Analysis MMLU-Pro2026 · 8 results · Reference
Graduate-Level Google-Proof Q&A2023 · 140 results · Reference
SuperGPQA: Scaling LLM Evaluation Across 285 Graduate Disciplines2025 · 38 results · Reference
Massive Multitask Language Understanding Professional2024 · 85 results · Reference
AGIEval2026 · 6 results · Reference
