BlogLLM Benchmark Scores as Published, Side by Side
IMO-AnswerBench: Results as Published
Every result the catalog holds for IMO-AnswerBench: 35 models from 10 developers, read from 12 pages. Each number is as printed by the company named beside it; nothing is ranked or rescored. Models are listed by developer, newest first.
Area Science and MathMaintainer Google DeepMind (Thang Luong and co-authors)Introduced 2025Data version 2026103001, as of 2026-10-30
IMO-AnswerBench is the short-answer part of IMO-Bench: 400 Olympiad-level problems whose answers can be checked automatically. The score is the share of problems answered correctly.
IMO-AnswerBench
| Model | Value | Printed by | Setup as printed |
|---|---|---|---|
| GPT-5.4OpenAI, released 2026-03-05 | 91.4% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15; printed on 2 more pages | Effort: xhigh |
| GPT-5.2OpenAI, released 2025-12-11 | 86.3% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 2 more pages | - |
| GPT-5OpenAI, released 2025-08-07 | 76% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15; printed on 1 more page | Effort: high; Tools: none |
| Claude Opus 4.8Anthropic, released 2026-05-28 | 83.5% | Z.ai, GLM-5.2 model cardchecked 2026-10-15 | - |
| Claude Opus 4.6Anthropic, released 2026-02-05 | 75.3% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15; printed on 2 more pages | Effort: max |
| Claude Opus 4.5Anthropic, released 2025-11-24 | 84% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 3 more pages | - |
| Claude Sonnet 4.5Anthropic, released 2025-09-29 | 65.9% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15; printed on 1 more page | Effort: thinking; Tools: none |
| Gemma 4 26B A4BGoogle, released 2026-04-02 | 74.3% | Alibaba, Qwen3.6-35B-A3B model cardchecked 2026-10-15 | - |
| Gemma 4 31BGoogle, released 2026-04-02 | 74.5% | Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more page | - |
| Gemini 3.1 Pro PreviewGoogle, released 2026-02-19 | 81% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15; printed on 3 more pages | Effort: high |
| Gemini 3 ProGoogle, released 2025-11-18 | 83.3% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 3 more pages | - |
| Grok 4xAI, released 2025-07-09 | 73.1% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15 | Tools: none |
| DeepSeek-V4-Flash (Preview)DeepSeek, released 2026-04-24 | 88.4% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15; printed on 1 more page | Effort: max |
| 41.9% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15 | Effort: non-think | |
| 85.1% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15 | Effort: high | |
| DeepSeek-V4-Pro (Preview)DeepSeek, released 2026-04-24 | 88% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15; printed on 2 more pages | Effort: high |
| 35.3% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15 | Effort: non-think | |
| 89.8% | DeepSeek, DeepSeek-V4-Pro (preview) model cardchecked 2026-10-15 | Effort: max | |
| DeepSeek-V3.2DeepSeek, released 2025-12-01 | 78.3% | Moonshot AI, Kimi K2.5 model cardchecked 2026-10-15; printed on 3 more pages | - |
| DeepSeek-V3.2-ExpDeepSeek, released 2025-09-29 | 76% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15 | Tools: none |
| Qwen3.6-27BAlibaba, released 2026-04-22 | 80.8% | Alibaba, Qwen3.6-27B model cardchecked 2026-10-15 | - |
| Qwen3.6-35B-A3BAlibaba, released 2026-04-16 | 78.9% | Alibaba, Qwen3.6-35B-A3B model cardchecked 2026-10-15; printed on 1 more page | - |
| Qwen3.6-PlusAlibaba, released 2026-04-02 | 83.8% | Z.ai, GLM-5.1 model cardchecked 2026-10-15 | - |
| Qwen3.5-27BAlibaba, released 2026-02-24 | 79.9% | Alibaba, Qwen3.6-27B model cardchecked 2026-10-15; printed on 1 more page | - |
| Qwen3.5-35B-A3BAlibaba, released 2026-02-24 | 76.8% | Alibaba, Qwen3.6-35B-A3B model cardchecked 2026-10-15 | - |
| Qwen3.5-397B-A17BAlibaba, released 2026-02-16 | 80.9% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15; printed on 2 more pages | - |
| Qwen3-Max-ThinkingAlibaba | 83.9% | Alibaba, Qwen3.5-397B-A17B model cardchecked 2026-10-15 | - |
| Qwen3.7-MaxAlibaba | 90% | Z.ai, GLM-5.2 model cardchecked 2026-10-15 | - |
| Kimi K2.6Moonshot AI, released 2026-04-20 | 86% | Moonshot AI, Kimi K2.6 model cardchecked 2026-10-15; printed on 2 more pages | - |
| Kimi K2.5Moonshot AI, released 2026-01-27 | 81.8% | Moonshot AI, Kimi K2.5 model cardchecked 2026-10-15; printed on 4 more pages | - |
| Kimi K2 ThinkingMoonshot AI, released 2025-11-06 | 78.6% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15; printed on 1 more page | Effort: thinking; Tools: none |
| Kimi K2 Instruct 0905Moonshot AI, released 2025-09-05 | 45.8% | Moonshot AI, Kimi K2 Thinking release pagechecked 2026-10-15 | Tools: none |
| GLM-5.2Z.ai, released 2026-06-16 | 91% | Z.ai, GLM-5.2 model cardchecked 2026-10-15 | - |
| GLM-5.1Z.ai, released 2026-04-07 | 83.8% | Z.ai, GLM-5.1 model cardchecked 2026-10-15; printed on 3 more pages | - |
| GLM-5Z.ai, released 2026-02-12 | 82.5% | Z.ai, GLM-5 model cardchecked 2026-10-15; printed on 1 more page | - |
| GLM-4.7Z.ai, released 2025-12-22 | 82% | Z.ai, GLM-4.7 model cardchecked 2026-10-15; printed on 1 more page | - |
| GLM-4.6Z.ai, released 2025-09-30 | 73.5% | Z.ai, GLM-4.7 model cardchecked 2026-10-15 | - |
| MiniMax M2.7MiniMax, released 2026-03-18 | 68.3% | NVIDIA, NVIDIA Nemotron 3 Ultra 550B-A55B model cardchecked 2026-10-15; printed on 1 more page | Tools: none |
| 75.1% | NVIDIA, NVIDIA Nemotron 3 Ultra 550B-A55B model cardchecked 2026-10-15 | Tools: with tools | |
| Nemotron 3 UltraNVIDIA, released 2026-06-04 | 92.3% | NVIDIA, NVIDIA Nemotron 3 Ultra 550B-A55B model cardchecked 2026-10-15 | Tools: with tools |
| 88.6% | NVIDIA, NVIDIA Nemotron 3 Ultra 550B-A55B model cardchecked 2026-10-15 | Tools: none |
Every number here is read from the same catalog file as the benchmark table, data version 2026103001, and is the number the table shows for that model and test. Where one page printed a model under several setups, each value is listed. The catalog is public in armanobosyan/sugra-llm-benchmarks; a wrong number can be reported from the table, with Report an error under the value, or as an issue in that repository.