GigaChat 2 Pro: benchmark results
Sber's GigaChat 2 Pro, the mid tier in their three-tier GigaChat 2.0 line (Lite, Pro, Max), with a 128K context window and up to 8K output tokens for everyday tasks. Provider: Sber. Released 2025-03-13. Access: API.
Aggregate Rating 1531 ± 30, rank #676 of 1635 rated models, from 29 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| MERA - ruHumanEval | 62.74 | pass@1 (%) | 91.4 |
| MERA - PARus | 93.4 | Accuracy (%) | 90.5 |
| MERA - ruCodeEval | 59.88 | pass@1 (%) | 90 |
| MERA - ruWorldTree | 99.05 | Accuracy (%) | 88.8 |
| MERA - USE | 53.43 | Grade, normalized (%) | 88.6 |
| MERA - ruTiE | 88.31 | Accuracy (%) | 88.6 |
| MERA - ruModAr | 94.33 | EM (%) | 85.8 |
| MERA - RWSD | 66.54 | Accuracy (%) | 84.3 |
| MERA - ruOpenBookQA | 93 | Accuracy (%) | 81.2 |
| MERA | 0.65 | Total score | 80.5 |
| MERA - MathLogicQA | 77.52 | Accuracy (%) | 78.8 |
| MERA - MaMuRAMu | 83.07 | Accuracy (%) | 77.6 |
Interactive version: theaggregate.ai/model?slug=gigachat-2-pro · How It Works · Data refreshed daily, snapshot 2026-11-05.