deepseek-llm-67B-base: benchmark results
DeepSeek LLM 67B base checkpoint. Provider: DeepSeek. Released 2023-11-29. Access: Open.
Aggregate Rating 1454 ± 19, rank #973 of 1656 rated models, from 20 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Chinese LLM - HellaSwag | 72.83 | Accuracy (%) | 96.7 |
| Open LLM Leaderboard v1 - MMLU | 71.78 | Accuracy (%) (5-shot) | 95.2 |
| Open LLM Leaderboard v1 - WinoGrande | 84.14 | Accuracy (%) (5-shot) | 92.1 |
| Open Chinese LLM - ARC Challenge | 61.86 | Accuracy (%) | 91.4 |
| Open Chinese LLM - WinoGrande | 69.06 | Accuracy (%) | 89.1 |
| Open LLM Leaderboard v1 - HellaSwag | 87.1 | Normalized accuracy (%) (10-shot) | 84.5 |
| Open Chinese LLM Leaderboard | 64.38 | Official Nonzero-Task Mean (%) | 79.9 |
| Open Chinese LLM - CMMLU | 66.86 | Accuracy (%) | 77.5 |
| TheoremQA | 25.3 | Score (%) | 75.8 |
| Open Chinese LLM - C-Eval Semantic | 81.81 | Accuracy (%) | 71 |
| Open LLM Leaderboard v1 - ARC Challenge | 65.44 | Normalized accuracy (%) (25-shot) | 70 |
| Open LLM Leaderboard v1 - GSM8K | 56.71 | Accuracy (%) (5-shot) | 68.5 |
Interactive version: theaggregate.ai/model?slug=deepseek-llm-67b-base · How It Works · Data refreshed daily, snapshot 2026-10-20.