openbuddy-llama3-8B-v21.1-8k: benchmark results
OpenBuddy's multilingual chatbot fine-tune of Llama 3 8B with an 8k context, tuned for English and Chinese conversation. Provider: OpenBuddy. Released 2024-04-20. Access: Open.
Aggregate Rating 1417 ± 16, rank #1159 of 1656 rated models, from 28 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open CoT - LSAT Analytical Reasoning | 7.83 | CoT Gain (%) | 86.3 |
| Open Korean LLM - Ko-IFEval | 40.7 | Strict accuracy, prompt/instruction mean (%) | 83.4 |
| Open CoT - LSAT Reading Comprehension | 18.96 | CoT Gain (%) | 82.1 |
| Open Korean LLM - KorNAT-Harmless | 64.82 | Normalized accuracy (%) | 77.2 |
| Open Korean LLM - Ko-GSM8K (flexible extract) | 42 | Exact match, flexible extract (%) | 77 |
| Open Korean LLM - Ko-EQ-Bench | 28.69 | EQ-Bench score | 76.5 |
| Open LLM Leaderboard v1 - GSM8K | 61.18 | Accuracy (%) (5-shot) | 73.6 |
| Open CoT Leaderboard | 10.81 | Average CoT Gain (%) | 71 |
| Open Korean LLM - Ko-GPQA-Diamond | 24.75 | Normalized accuracy (%) | 69.5 |
| Open CoT - LSAT Logical Reasoning | 14.12 | CoT Gain (%) | 68.3 |
| Open LLM Leaderboard - IFEval | 55.7 | Score | 67.5 |
| Open Korean LLM Leaderboard | 38.47 | Average Score (%) | 67.3 |
Interactive version: theaggregate.ai/model?slug=openbuddy-llama3-8b-v21-1-8k · How It Works · Data refreshed daily, snapshot 2026-10-20.