MERA - ruWorldTree: leaderboard
MERA is the Russian open benchmark run by the AI Alliance Russia; this is its ruWorldTree task: elementary science questions with an explanation chain, in Russian. Each MERA task keeps its own native metric, so the boards are not on one shared scale.
Metric: Accuracy (%). Source: mera.a-ai.ru. Saturation forecast: Estimated already saturated. 632 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.2 | 99.62 | #110 |
| 2 | Claude Opus 4.6 | 99.62 | #63 |
| 3 | Qwen 3.5 27B | 99.62 | #187 |
| 4 | GLM-5.1 | 99.62 | #97 |
| 5 | Claude Opus 4.5 | 99.62 | #83 |
| 6 | GigaChat 2 Max | 99.62 | #536 |
| 7 | Gemini 2.0 Flash | 99.43 | #342 |
| 8 | Claude 3.7 Sonnet | 99.43 | #249 |
| 9 | Qwen 3.5 122B A10B | 99.43 | #173 |
| 10 | Qwen 3 235B A22B 2507 (Thinking) | 99.43 | #260 (Qwen 3 235B A22B 2507) |
| 11 | GPT-5.4 | 99.24 | #79 |
| 12 | Qwen 2.5 72B Instruct | 99.24 | #454 |
| 13 | Kimi K2.5 | 99.24 | #144 |
| 14 | Gemma 4 31B (IT) | 99.24 | #176 |
| 15 | Qwen 3.6 35B A3B | 99.24 | #226 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mera-ruworldtree · How It Works · Data refreshed daily, snapshot 2026-10-20.