deepseek-llm-7B-base Benchmark Results

deepseek-llm-7B-base: benchmark results

DeepSeek LLM 7B base checkpoint. Provider: DeepSeek. Released 2023-11-29. Access: Open.

Aggregate Rating 1319 ± 22, rank #1443 of 1656 rated models, from 25 benchmark results.

Benchmark	Score	Metric	Percentile
Open Chinese LLM - HellaSwag	61.91	Accuracy (%)	65.1
Open Chinese LLM - WinoGrande	64.01	Accuracy (%)	63.2
Open Portuguese LLM - OAB Exams	35.67	Accuracy (%)	33
Open Portuguese LLM - BLUEX	38.39	Accuracy (%)	31.7
Open Portuguese LLM - ENEM	41.78	Accuracy (%)	30.5
Open LLM Leaderboard - GPQA	27.35	Score	28.5
TheoremQA	15.7	Score (%)	27.3
Open Chinese LLM - ARC Challenge	47.1	Accuracy (%)	26.2
Open LLM Leaderboard - MuSR	37.38	Score	25.7
pfgen-bench - Completion Mode - Fluency	0.42	Fluency Score	20.9
Open LLM Leaderboard - MMLU-Pro	18.06	Score	19.6
Open LLM Leaderboard - BBH	35.03	Score	19.2

Interactive version: theaggregate.ai/model?slug=deepseek-llm-7b-base · How It Works · Data refreshed daily, snapshot 2026-10-20.