MERA - RWSD: leaderboard
MERA is the Russian open benchmark run by the AI Alliance Russia; this is its RWSD task: the Russian Winograd Schema Challenge, pronoun resolution that turns on world knowledge. 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: Around January 2027. 632 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.4 | 95 | #79 |
| 2 | Claude Opus 4.6 | 93.08 | #63 |
| 3 | Qwen 3.5 122B A10B | 90.77 | #173 |
| 4 | Qwen 3.5 27B | 90 | #187 |
| 5 | GLM-5.1 | 90 | #97 |
| 6 | Kimi K2.5 | 87.69 | #144 |
| 7 | Claude Opus 4.5 | 86.15 | #83 |
| 8 | GPT-5.2 | 84.23 | #110 |
| 9 | Qwen 3 235B A22B 2507 (Thinking) | 84.23 | #260 (Qwen 3 235B A22B 2507) |
| 10 | Qwen 3.6 27B | 83.85 | #170 |
| 11 | Qwen 3.5 35B A3B | 82.31 | #256 |
| 12 | Claude 3.7 Sonnet | 78.85 | #249 |
| 13 | Qwen 3.6 35B A3B | 78.08 | #226 |
| 14 | Qwen 3 Next 80B A3B (Thinking) | 76.54 | #320 (Qwen 3 Next 80B A3B) |
| 15 | Gemma 4 31B (IT) | 76.15 | #176 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mera-rwsd · How It Works · Data refreshed daily, snapshot 2026-10-20.