LinAlg-Bench: leaderboard
We introduce LinAlg-Bench, a diagnostic benchmark evaluating 10 frontier large language models on structured linear algebra computation across a strict dimensional gradient of 3x3, 4x4, and 5x5 matrices. Spanning 9 task types and 660 SymPy-certified problems, the benchmark exhaustively evaluates 6,600 model outputs. Beyond binary accuracy, LinAlg-Bench introduces a three-stage automated forensic p
Metric: 3×3 Overall Model Accuracy (%) (self-reported). Source: benchmarklist.com. Saturation forecast: Around April 2027. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Pro | 100 | #80 |
| 2 | O1 | 100 | #183 |
| 3 | DeepSeek V3 | 99.5 | #321 |
| 4 | GPT-5.2 | 98.2 | #110 |
| 5 | Mistral Large | 98.2 | #710 |
| 6 | Claude Sonnet 4.5 | 95.9 | #142 |
| 7 | Qwen 3 235B A22B | 92.7 | #313 |
| 8 | Qwen 2.5 72B | 91.8 | #499 |
| 9 | GPT-4o | 90.9 | #343 |
| 10 | Llama 3.3 70B | 87.3 | #579 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=linalg-bench · How It Works · Data refreshed daily, snapshot 2026-10-20.