GeoAux-Bench (Core) - Olympiad Senior: leaderboard
Metric: Accuracy (%) on the 237 senior olympiad questions of GeoAux-Bench-Core (expert-annotated geometry problems needing auxiliary constructions; final answers checked by a Qwen3-30B-A3B-Thinking-2507 verifier; greedy decoding, 8,192 new tokens); higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 21 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 VL 235B A22B (Thinking) | 89.45 | #228 (Qwen 3 VL 235B A22B) |
| 2 | Qwen 3 VL 30B A3B (Thinking) | 83.97 | #338 (Qwen 3 VL 30B A3B) |
| 3 | Gemini 2.5 Pro | 82.28 | #145 |
| 4 | GPT-5 | 79.32 | #91 |
| 5 | Qwen 3 VL 235B A22B Instruct | 78.9 | #264 |
| 6 | Qwen 3 VL 30B A3B Instruct | 75.53 | #365 |
| 7 | Doubao-Seed-1.6 (Thinking) | 72.57 | |
| 8 | GLM-4.5V (Non-reasoning) | 72.15 | #339 (GLM-4.5V) |
| 9 | InternVL3-8B | 67.93 | #606 |
| 10 | Gemini 2.5 Flash | 64.56 | #237 |
| 11 | Claude Opus 4.1 (Thinking) | 58.23 | #132 (Claude Opus 4.1) |
| 12 | GPT-4o | 57.38 | #333 |
| 13 | GLM-4.1V-9B (Thinking) | 56.12 | #457 (GLM-4.1V-9B) |
| 14 | Claude Sonnet 4.5 (Thinking) | 55.7 | #138 (Claude Sonnet 4.5) |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=geoaux-bench-core-olympiad-senior · How It Works · Data refreshed daily, snapshot 2026-10-11.