MERA Code - ruHumanEval: leaderboard
MERA Code (Russian) ruHumanEval task: the Russian HumanEval port, scored by unit tests. Each task keeps its own metric — pass rates, CodeBLEU and chrF all appear across this suite — so read the metric label before comparing boards.
Metric: pass@1 (%). Source: mera.a-ai.ru. Saturation forecast: Not forecast. 66 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 Next 80B A3B (Thinking) | 82.26 | #320 (Qwen 3 Next 80B A3B) |
| 2 | Qwen 3 30B A3B 2507 (Thinking) | 79.88 | #376 (Qwen 3 30B A3B 2507) |
| 3 | Qwen 3 4B 2507 (Thinking) | 78.11 | #538 (Qwen 3 4B 2507) |
| 4 | DeepSeek V4 Pro | 77.87 | #102 |
| 5 | Qwen 3 14B | 77.5 | #537 |
| 6 | Qwen 3 32B | 76.59 | #443 |
| 7 | GLM-5.1 | 74.7 | #97 |
| 8 | Qwen 3 8B | 73.72 | #674 |
| 9 | Qwen 3 1.7B | 66.59 | #1196 |
| 10 | GPT-5.1 Codex Max | 66.1 | #100 |
| 11 | Claude Opus 4.6 | 65.85 | #63 |
| 12 | Qwen 3 4B | 65.73 | #829 |
| 13 | MiniMax-M2.7 | 61.71 | #258 |
| 14 | Gemini 2.5 Flash | 60.37 | #245 |
| 15 | GigaChat 2 Max | 52.44 | #536 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mera-code-ruhumaneval · How It Works · Data refreshed daily, snapshot 2026-10-20.