IOL-AI Challenge 2026 - Exact Match: leaderboard
Exact-match share of answers on the IOL 2026 Individual Contest problems, the strict half of the IOL-AI Challenge score with no credit for near-miss spellings. Higher is better. Source: arxiv.org/abs/2608.18011.
Metric: Exact Match (%; IOL 2026 Individual Contest). Source: arxiv.org. Saturation forecast: Around February 2027. 18 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.8 | 68.1 | #57 |
| 2 | Gemini 3.6 Flash | 45 | #74 |
| 3 | GPT-5.6 Sol | 43.1 | #29 |
| 4 | Kimi K2.6 | 35.8 | #176 |
| 5 | GLM-5.2 | 35 | #138 |
| 6 | DeepSeek V4 Pro | 31.5 | #188 |
| 7 | Qwen 3.8 27B | 30.9 | #206 |
| 8 | Muse Glimmer 30B (High) | 30.3 | #381 (Muse Glimmer 30B) |
| 9 | Qwen 3.7 Max | 29.4 | #143 |
| 10 | Inkling | 24.3 | #271 |
| 11 | Nemotron 3 Ultra | 20.1 | #427 |
| 12 | Qwen 3.6 27B (Thinking) | 19.2 | #345 (Qwen 3.6 27B) |
| 13 | Qwen 3.6 27B (Non-reasoning) | 14.9 | #345 (Qwen 3.6 27B) |
| 14 | Gemma 4 31B | 14.6 | #392 |
| 15 | Command A+ | 11.1 | #940 |
No result here: #4 Claude Opus 5.5, #5 GPT-6 Astra, #9 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=iol-ai-challenge-2026-exact-match · How It Works · Data refreshed daily, snapshot 2026-11-08.