Bielik-4.5B-v3.0-Instruct: benchmark results
Provider: SpeakLeash. Released 2025-07-30. Access: Open.
Aggregate Rating 1488 ± 18, rank #2538 of 10916 rated models, from 304 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open PL LLM - PolQA Open Book (generative, 5-shot) | 88.19 | Levenshtein Similarity (%) | 83.3 |
| Open PL LLM - CBD (multiple choice, 5-shot) | 35.32 | Macro F1 (%) | 81.9 |
| EuroEval French NLU - ScaLA FR | 43.78 | Matthews correlation coefficient (%) | 80.1 |
| Open PL LLM - DYK (generative, 5-shot) | 66.48 | Binary F1 (%) | 79.2 |
| Open PL LLM - PolQA Closed Book (generative, 5-shot) | 72.48 | Levenshtein Similarity (%) | 79.2 |
| Open PL LLM - PSC (multiple choice, 5-shot) | 83.97 | Binary F1 (%) | 77.7 |
| Polish Knowledge Benchmark - Theatre | 76.54 | Accuracy (%) | 77.2 |
| Open PL LLM - PPC (multiple choice, 5-shot) | 77.1 | Accuracy (%) | 75.4 |
| EuroEval French NLU - Allocine | 94.16 | Matthews correlation coefficient (%) | 74.6 |
| Open PL LLM - CBD (generative, 5-shot) | 32.43 | Macro F1 (%) | 74.5 |
| Open PL LLM - RAG | 67.98 | Average RAG Score (%) | 72.4 |
| EuroEval English NLU - SQuAD | 82.8 | Character F1 (%) | 72.1 |
Interactive version: theaggregate.ai/model?slug=bielik-4-5b-v3-0-instruct · How It Works · Data refreshed daily, snapshot 2026-11-08.