DeepSeek V3.2 (High) — benchmark results

DeepSeek V3.2 evaluated at the high reasoning-effort setting. Provider: DeepSeek. Released 2025-12-01. Access: Open.

Unified ELO 1623 ± 91, rank #431 of 1776 rated models, from 7 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
NonoBench36.7Overall Accuracy (%)69.1
EnterpriseOps-Gym23.8Task Success Rate (self-reported)47.6
IOI10.67Score (self-reported)43.4
Vals AI CaseLaw v255.41Accuracy (%)38.5
FutureEval7.74Unified Forecasting Score37.8
Vals AI ProofBench4Accuracy (%)8.4
Finance Agent v1.12.35Score (self-reported)3.6

Interactive version: theaggregate.ai/model?slug=deepseek-v3-2-high · How the rankings work · Data refreshed daily, snapshot 2026-07-22.