DeepSeek V2 — benchmark results

DeepSeek's 236B open MoE (21B active) with Multi-head Latent Attention and a 128K context. Provider: DeepSeek. Released 2024-05-06. Access: Open.

Unified ELO 1442 ± 22, rank #1037 of 1776 rated models, from 11 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
ARC Challenge (AI2)92.2Accuracy (%)94.9
WinoGrande86.3Accuracy (%)92.5
HellaSwag87.1Accuracy (%)92.1
Big-Bench Hard78.8Average (%)87.5
PIQA83.9Accuracy (%)86.7
MMLU78.4Accuracy (%)77
MixEval51.7Score68.6
TriviaQA80Accuracy (%)64.1
LLM2014 Logic 2024-0539.45Score (%)57.1
LLM2014 Logic 2024-0639.49Score (%)41.4
Epoch AI - ECI124.57ECI Score21.4

Interactive version: theaggregate.ai/model?slug=deepseek-v2 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.