Claude Sonnet 4vsDeepSeek-V4-Pro

Claude Sonnet 4
DeepSeek-V4-Pro
API pricing
Cheapest input
$3.00Amazon Bedrock$0.7078StreamLake
Cheapest output
$15.00Amazon Bedrock$1.4157StreamLake
Benchmarks
BullshitBench v2
30%14%
GPQA Diamond
75.4%90.1%
Benchmarks
SWE-Bench Verified
72.7%
LiveCodeBench
93.5%
BrowseComp
83.4%
Overview
CompanyAnthropicDeepSeek
Release dateMay 22 2025Apr 24 2026
AccessProprietaryOpen Weight

Other comparisons

Claude Sonnet 4vsGPT-6 AstraDeepSeek-V4-ProvsGPT-6 AstraClaude Sonnet 4vsGemini 3.8 FlashDeepSeek-V4-ProvsGemini 3.8 FlashClaude Sonnet 4vsMuse Spark 1.3DeepSeek-V4-ProvsMuse Spark 1.3Claude Sonnet 4vsGrok 4.6DeepSeek-V4-ProvsGrok 4.6Claude Sonnet 4vsMistral Medium 3.5DeepSeek-V4-ProvsMistral Medium 3.5Claude Sonnet 4vsKimi K3DeepSeek-V4-ProvsKimi K3

Frequently asked questions

Claude Sonnet 4 and DeepSeek-V4-Pro are evenly matched across the 2 benchmarks they both report (BullshitBench v2, GPQA Diamond). Claude Sonnet 4 shipped 337 days before DeepSeek-V4-Pro, so benchmark comparisons should account for the intervening progress.

Claude Sonnet 4 is proprietary, while DeepSeek-V4-Pro is open weight.

On BullshitBench v2, Claude Sonnet 4 leads at 30% vs DeepSeek-V4-Pro at 14%. On GPQA Diamond, DeepSeek-V4-Pro leads at 90.1% vs Claude Sonnet 4 at 75.4%.