DeepSeek13 Aug 2026Model

DeepSeek-V4-Pro-0813

Benchmark results

Benchmark	Score	Source check
Humanity's Last ExamNo tools	42.7%	Corroborated The same number is printed in 2 other lab documents: 1 2
Humanity's Last ExamWith tools	60.0%	Corroborated The same number is printed in 2 other lab documents: 1 2
GPQA Diamondreported later, in DeepSeek-V4.1-Flash: Smarter, Faster, More Efficient	92.4%	Corroborated The same number is printed in 1 other lab document: 1
AutomationBenchpublic set	31.8%	Checked A person checked this number against its source on 20 Oct 2026. This confirms the number matches what the lab published, not that the score was reproduced.
set not stated · reported later, in DeepSeek-V4.1-Flash: Smarter, Faster, More Efficient	43.2%	Corroborated The same number is printed in 1 other lab document: 1
DeepSWE v1.1	62.7%	Corroborated The same number is printed in 2 other lab documents: 1 2

Sources

Reported by DeepSeek