WizardLM-2 8x22B: benchmark results
Microsoft's Mixtral-8x22B finetune briefly pulled days after its April 2024 release over missing toxicity testing, surviving via community mirrors. Provider: Microsoft. Released 2024-04-15. Access: Open.
Aggregate Rating 1523 ± 13, rank #703 of 1656 rated models, from 95 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open PL LLM - EQ-Bench | 62.81 | EQ-Bench Score | 97 |
| Open PL LLM - KLEJ NER (generative, 5-shot) | 59.33 | Accuracy (%) | 95.9 |
| Open LLM Leaderboard - GPQA | 38.17 | Score | 95.3 |
| Open Portuguese LLM - ASSIN2 RTE | 94.12 | Macro F1 (%) | 95 |
| Open PL LLM - CBD (multiple choice, 5-shot) | 38.98 | Macro F1 (%) | 94.1 |
| Open Portuguese LLM - OAB Exams | 56.45 | Accuracy (%) | 89.9 |
| Open Portuguese LLM - ENEM | 75.86 | Accuracy (%) | 89.7 |
| LogicKor - Reasoning | 8.5 | Score (0-10) | 89.1 |
| Open Chinese LLM - GSM8K | 64.82 | Accuracy (%) | 89.1 |
| Open PL LLM - 8TAGS (generative, 5-shot) | 78.29 | Accuracy (%) | 89 |
| Open PL LLM - KLEJ NER (multiple choice, 5-shot) | 56.75 | Accuracy (%) | 88.8 |
| Open LLM Leaderboard - BBH | 63.77 | Score | 88 |
Interactive version: theaggregate.ai/model?slug=wizardlm-2-8x22b · How It Works · Data refreshed daily, snapshot 2026-10-20.