Vals AI MGSM: leaderboard
MGSM multilingual grade-school math run independently by Vals AI on frontier models; accuracy (%).
Metric: Accuracy (%). Source: www.vals.ai. Source: benchmarklist.com. Saturation forecast: Estimated already saturated. 75 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.5 (20251101) (Thinking) | 95.2 | #74 (Claude Opus 4.5 (20251101)) |
| 2 | Claude Opus 4.5 (20251101) | 94.76 | #74 |
| 3 | Claude Opus 4.1 (20250805) (Thinking) | 94.44 | #132 (Claude Opus 4.1 (20250805)) |
| 4 | Claude Sonnet 4.5 (Thinking) | 94.33 | #142 (Claude Sonnet 4.5) |
| 5 | Claude Opus 4.1 (20250805) | 94.22 | #132 |
| 6 | GPT-5.2 | 94 | #110 |
| 7 | GPT-5.2 (2025-12-11) (xHigh) | 94 | |
| 8 | Gemini 3 Pro (Preview) (High) | 93.93 | #66 (Gemini 3 Pro (Preview)) |
| 9 | Claude Opus 4 (20250514) | 93.78 | #154 |
| 10 | Claude Opus 4 | 93.78 | #163 |
| 11 | O4 Mini (2025-04-16) (High) | 93.42 | #179 (O4 Mini (2025-04-16)) |
| 12 | Gemini 3 Flash (Preview) (High) | 93.31 | #82 (Gemini 3 Flash (Preview)) |
| 13 | Claude Sonnet 4 (20250514) | 93.02 | #213 |
| 14 | Claude 3.7 Sonnet (20250219) (Thinking) | 92.98 | #201 (Claude 3.7 Sonnet (20250219)) |
| 15 | GPT-5.1 (2025-11-13) (High) | 92.98 | #101 (GPT-5.1 (2025-11-13)) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=vals-ai-mgsm · How It Works · Data refreshed daily, snapshot 2026-10-20.