DeepSeek V4.1 Flash (Non-reasoning): benchmark results

Provider: DeepSeek. Released 2026-09-10. Access: Open.

Unified ELO 1623 ± 1, rank #341 of 1935 rated models, from 63 benchmark results.

Strongest benchmark results

BenchmarkScoreMetricPercentile
AA-Omniscience Index - Software Engineering (SWE) - Go30Omniscience Index81.7
AA-Omniscience Hallucination Rate53.99Hallucination Rate (%)78.6
ParseBench - Semantic Formatting66.05Semantic Formatting Score77.9
Artificial Analysis Intelligence Index24.67Intelligence Index77.3
AA-Omniscience Index - Business-6Omniscience Index76.6
AA-Omniscience Index - Software Engineering (SWE) - Python13.5Omniscience Index75.4
AA-Omniscience Index - Software Engineering (SWE) - JavaScript15.45Omniscience Index74.5
AutomationBench-AA - Finance - Raw Objective Completion79.68Raw Objective Completion (%)73.7
AA GDPval1327.69ELO72.8
AA-Omniscience Index - Humanities & Social Sciences-7.9Omniscience Index72.6
AA-Omniscience Index - Software Engineering (SWE) - HTML20Omniscience Index72.3
AA-Omniscience Index - Law-12.2Omniscience Index71.7

Interactive version: theaggregate.ai/model?slug=deepseek-v4-1-flash-non-reasoning · How It Works · Data refreshed daily, snapshot 2026-09-25.