openbuddy-llama3-70B-v21.2-32k: benchmark results
OpenBuddy's multilingual chatbot fine-tune of Llama 3 70B Instruct with a 32k context, aimed at English and Chinese use. Provider: OpenBuddy. Released 2024-06-12. Access: Open.
Aggregate Rating 1595 ± 44, rank #1056 of 10916 rated models, from 14 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Chinese LLM - CMMLU | 71.61 | Accuracy (%) | 92.5 |
| Open Chinese LLM - GSM8K | 67.63 | Accuracy (%) | 91.6 |
| Open Chinese LLM Leaderboard | 68.9 | Official Nonzero-Task Mean (%) | 91.4 |
| Open LLM Leaderboard - BBH | 65.07 | Score | 91.4 |
| Open LLM Leaderboard - MuSR | 45.8 | Score | 87 |
| Open Chinese LLM - C-Eval Semantic | 87.87 | Accuracy (%) | 86.7 |
| Open LLM Leaderboard - MMLU-Pro | 48.32 | Score | 86.7 |
| Open LLM Leaderboard - GPQA | 34.23 | Score | 85.1 |
| Open LLM Leaderboard - IFEval | 70.1 | Score | 83.1 |
| Open Chinese LLM - HellaSwag | 66.73 | Accuracy (%) | 81.7 |
| Open Chinese LLM - ARC Challenge | 57.68 | Accuracy (%) | 76 |
| Open Chinese LLM - WinoGrande | 65.98 | Accuracy (%) | 75.7 |
Interactive version: theaggregate.ai/model?slug=openbuddy-llama3-70b-v21-2-32k · How It Works · Data refreshed daily, snapshot 2026-11-08.