Open PL LLM - PPC (generative, 5-shot): leaderboard
Polish Paraphrase Corpus: classifying sentence pairs as exact paraphrases, close in meaning or unrelated. Open PL LLM Leaderboard (SpeakLeash), 5-shot, generative variant: the answer is extracted from generated text; accuracy (%), higher is better. Runs not available on the Hugging Face Hub are excluded.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 320 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 32B | 83.6 | #544 |
| 2 | Llama 3.1 405B Instruct FP8 | 82.5 | #400 |
| 3 | QwQ 32B-Preview | 82.2 | #671 |
| 4 | Qwen 2.5 72B Instruct | 81.8 | #469 |
| 5 | QwQ-32B | 81.5 | #422 |
| 6 | Qwen 2.5 14B | 81.5 | #716 |
| 7 | Qwen 2 72B | 81.4 | #515 |
| 8 | Qwen 2 72B Instruct | 81.1 | #579 |
| 9 | Qwen 2.5 72B | 81.1 | #512 |
| 10 | aya-expanse-32B | 80.8 | #687 |
| 11 | cogito-v1-preview-qwen-14B | 80.7 | |
| 12 | Qwen 2.5 32B Instruct | 80.6 | #509 |
| 13 | Llama 3.3 70B Instruct | 79.9 | #541 |
| 14 | Qwen 3 30B A3B | 79.6 | #496 |
| 15 | Bielik-11B-v2.1-Instruct | 79.3 | #679 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-pl-llm-ppc-generative-5-shot · How It Works · Data refreshed daily, snapshot 2026-11-05.