MERA - RCB: leaderboard
MERA is the Russian open benchmark run by the AI Alliance Russia; this is its RCB task: the Russian Commitment Bank, a textual-entailment task over embedded clauses. 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 2029. 632 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 63.24 | #63 |
| 2 | GLM-5.1 | 63.24 | #97 |
| 3 | Gemma 4 31B (IT) | 61.42 | #176 |
| 4 | Qwen 3 235B A22B 2507 (Thinking) | 61.19 | #260 (Qwen 3 235B A22B 2507) |
| 5 | Mixtral 8x22B Instruct (v0.1) | 61.19 | #756 |
| 6 | Qwen 3.5 27B | 60.5 | #187 |
| 7 | Llama 3.1 405B Instruct | 60.5 | #457 |
| 8 | Qwen 3.6 27B | 60.5 | #170 |
| 9 | Kimi K2.5 | 60.27 | #144 |
| 10 | Qwen 3.5 122B A10B | 60.27 | #173 |
| 11 | Qwen 3.6 35B A3B | 60.05 | #226 |
| 12 | Qwen 3 30B A3B 2507 (Thinking) | 60.05 | #376 (Qwen 3 30B A3B 2507) |
| 13 | Llama 3.1 70B Instruct | 59.82 | #561 |
| 14 | Gemini 1.5 Pro | 59.82 | #445 |
| 15 | Claude Opus 4.5 | 59.82 | #83 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mera-rcb · How It Works · Data refreshed daily, snapshot 2026-10-20.