Note: Overall leaderboard rankings may not reflect true model quality — individual benchmarks give a clearer picture. ARC-Challenge MMLU GPQA GSM8K Artificial Analysis Intelligence Index v4.1
← Back to leaderboard

MBPP+

27 models

Top 5 Models Performance

orcarouter: orcacyber-nano-1.0 ######################################## 84.66
Sber: gigachat3.5-432b-a28b-base ####################################### 83.33
Prism ML: ternary-bonsai-2-27b ####################################### 83.07
Prism ML: ternary-bonsai-27b ###################################### 81.22
webai-official: twil-8b ##################################### 79.1
– 970 – 1.1T
– 2019 – 2026
Rank Model Score
❓ 1 orcarouter: orcacyber-nano-1.0 84.66
🇷🇺 2 Sber: gigachat3.5-432b-a28b-baseСбер 83.33
❓ 3 Prism ML: ternary-bonsai-2-27b 83.07
❓ 4 Prism ML: ternary-bonsai-27b 81.22
❓ 5 webai-official: twil-8b 79.1
🇨🇳 6 Tencent: hy3-preview-base腾讯控股有限公司 78.71
🇫🇷 7 Mistral AI: mistral-small-3.2-24b-instruct-2506 78.33
🇨🇳 8 Qwen: Qwen3 4B通义千问 77.6
❓ 9 essentialai: rnj-1-instruct 75.7
🇺🇸 10 OpenAI: gpt-oss-20b 75.6
❓ 11 zyphra: zaya1-base 75.4
🇨🇳 12 Xiaomi: MiMo-V2.5 Pro小米集团 74.1
🇨🇳 13 Qwen: qwen2.5-coder-7b-instruct通义千问 71.7
🇨🇳 14 Tencent: youtu-llm-2b腾讯控股有限公司 71.7
🇨🇳 15 Qwen: Qwen3 8B通义千问 71.2
❓ 16 essentialai: rnj-1 70.6
🇨🇳 17 Qwen: qwen3-1.7b通义千问 67.7
🇨🇳 18 Qwen: qwen3-8b-base通义千问 66.4
🇺🇸 19 Apple: diffucoder-7b-instruct 61.9
🇺🇸 20 Apple: diffucoder-7b-base 60.9
🇦🇪 21 Technology Innovation Institute: falcon-h1-1.5b-deep-base 60.32
🇺🇸 22 Hugging Face Smol Models Research: smollm3-3b 56.7
🇦🇪 23 Technology Innovation Institute: falcon-h1-1.5b-deep-instruct 56.61
🇦🇪 24 Technology Innovation Institute: falcon-h1-1.5b-instruct 56.35
🇦🇪 25 Technology Innovation Institute: falcon-h1-1.5b-base 55.03
🇺🇸 26 Arcee AI: trinity-nano-base 44.7
🇨🇳 27 DeepSeek: deepseek-r1-distill-qwen-1.5b深度求索 44.2