deepseek-llm-67B-base Benchmark Results

deepseek-llm-67B-base — benchmark results

DeepSeek LLM 67B base checkpoint. Provider: DeepSeek. Released 2023-11-29. Access: Open.

Unified ELO 1440 ± 18, rank #1041 of 1776 rated models, from 10 benchmark results.

Benchmark	Score	Metric	Percentile
Open Chinese LLM - HellaSwag	72.83	Accuracy (%)	96.7
Open Chinese LLM - ARC Challenge	61.86	Accuracy (%)	91.4
Open Chinese LLM - WinoGrande	69.06	Accuracy (%)	89
Open Chinese LLM Leaderboard	64.38	Average Score (%)	79.2
Open Chinese LLM - CMMLU	66.86	Accuracy (%)	77.4
Open Chinese LLM - C-Eval Semantic	81.81	Accuracy (%)	70.9
Open Chinese LLM - GSM8K	48.67	Accuracy (%)	62.6
InfiBench	39.87	Score (%)	54.3
Science Leaderboard	39.4	Average Accuracy (%)	42.4
Open Chinese LLM - TruthfulQA MC	49.6	Accuracy (%)	17.2

Interactive version: theaggregate.ai/model?slug=deepseek-llm-67b-base · How the rankings work · Data refreshed daily, snapshot 2026-07-22.