DeepSeek-V4.1-FlashvsComposer 2.5

DeepSeek-V4.1-Flash
Composer 2.5
Benchmarks
Terminal-Bench 2.1
90.6%73%
Benchmarks
SWE-Bench Pro
54%
SWE-Bench Multilingual
79.8%
DeepSWE 1.1
74.2%
DeepSWE 1.0
18%
Next.js Evals
81%
NL2Repo-Bench
65.4%
Terminal-Bench 4.0
31.2%
Terminal-Bench 3.0
30%
Terminal-Bench 2.0
69.3%
CyberGym
88.1%
Humanity's Last Exam · no tools
36.8%
Humanity's Last Exam · with tools
63.9%
GPQA Diamond
90.9%
Agent's Last Exam · pass@1
31.8%
AutomationBench
54.8%
Chartography · with tools
78.9%
threejseval
1120
Overview
CompanyDeepSeekSpaceXAI
Release dateSep 10 2026May 18 2026
AccessProprietaryProprietary

Other comparisons

DeepSeek-V4.1-FlashvsClaude Fable 5.1Composer 2.5vsClaude Fable 5.1DeepSeek-V4.1-FlashvsGPT-6 AstraComposer 2.5vsGPT-6 AstraDeepSeek-V4.1-FlashvsGemini 3.8 FlashComposer 2.5vsGemini 3.8 FlashDeepSeek-V4.1-FlashvsMuse Spark 1.3Composer 2.5vsMuse Spark 1.3DeepSeek-V4.1-FlashvsMistral Medium 3.5Composer 2.5vsMistral Medium 3.5DeepSeek-V4.1-FlashvsKimi K3Composer 2.5vsKimi K3

Frequently asked questions

DeepSeek-V4.1-Flash leads Composer 2.5 on 1 of the 1 benchmark they both report (Terminal-Bench 2.1). Composer 2.5 shipped 115 days before DeepSeek-V4.1-Flash, so benchmark comparisons should account for the intervening progress.

Published specifications for these two models are limited — see each model page for the latest details.

On Terminal-Bench 2.1, DeepSeek-V4.1-Flash leads at 90.6% vs Composer 2.5 at 73%.