Results

9 results

MATH-Vision Original results
ProviderModelScoreSourceRegistry No.
Moonshot AIKimi K3 (max)94.3%raw.githubusercontent.com for Kimi K3 (max) on MATH-Vision Original (opens in a new tab)AccuracyReported Technical reportMoonshot AISelf-reportedRecord for Kimi K3 on MATH-Vision Original120003
Google (DeepMind)Gemma 4 26B A4B Instruct (thinking mode)82.4%huggingface.co for Gemma 4 26B A4B Instruct (thinking mode) on MATH-Vision Original (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 26B A4B Instruct on MATH-Vision Original35001
Google (DeepMind)Gemma 4 31B Instruct (thinking mode)85.6%huggingface.co for Gemma 4 31B Instruct (thinking mode) on MATH-Vision Original (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 31B Instruct on MATH-Vision Original35002
Google (DeepMind)Gemma 4 E2B Instruct (thinking mode)52.4%huggingface.co for Gemma 4 E2B Instruct (thinking mode) on MATH-Vision Original (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 E2B Instruct on MATH-Vision Original35003
Google (DeepMind)Gemma 4 E4B Instruct (thinking mode)59.5%huggingface.co for Gemma 4 E4B Instruct (thinking mode) on MATH-Vision Original (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 E4B Instruct on MATH-Vision Original35004
Google (DeepMind)Gemma 4 12B Instruct (thinking mode)79.7%huggingface.co for Gemma 4 12B Instruct (thinking mode) on MATH-Vision Original (opens in a new tab)AccuracyReported Model cardGoogleSelf-reportedGoogle DeepMindRecord for Gemma 4 12B Instruct on MATH-Vision Original35005
Alibaba (Tongyi)Qwen3.6-Plus88.0%alibabacloud.com for Qwen3.6-Plus on MATH-Vision Original (opens in a new tab)AccuracyReported Launch postAlibaba Cloud CommunitySelf-reportedQwen TeamRecord for Qwen3.6-Plus on MATH-Vision Original130004
Alibaba (Tongyi)Qwen3.7-Plus90.3%alibabacloud.com for Qwen3.7-Plus on MATH-Vision Original (opens in a new tab)AccuracyReported Launch postAlibaba Cloud CommunitySelf-reportedQwen TeamRecord for Qwen3.7-Plus on MATH-Vision Original130007
Moonshot AIKimi K2.6 (thinking)87.4%huggingface.co for Kimi K2.6 (thinking) on MATH-Vision Original (opens in a new tab)AccuracyReported Model cardMoonshot AISelf-reportedRecord for Kimi K2.6 on MATH-Vision Original120002