Unverified50% confidenceBenchmarkExact time

DeepSeek的编程智能体DeepSWE在SWE-bench基准测试中从预览版的12.8分提升至62.7分

1

Sources

50%

Confidence

Medium-term (~90 days)

Relevance

8/20/2026

First Seen

Valid until: 11/18/2026

Sources

DeepSeek V4 Pro与Grok 4.6同日发布：AI大厂Agent之战全面打响

Related Claims

UnverifiedDeepSeek V4 Pro的软件工程能力从预览版的12.8分提升到62.7分，网络安全达到83.3分75% similar UnverifiedDeepSeek V4 Pro的代码工程62.7分比Fable 5低7.3分，数据科学难级67.2分落后Opus 4.8约4.5分，高难推理60分位居全球第二75% similar UnverifiedDeepSeek的代码能力在多项基准测试中达到国际领先水平74% similar UnverifiedDeepSeek作为免费模型，在AI编程工具测试中表现亮眼，超出预期74% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/778758

API

curl https://kongchang.com/api/v1/knowledge/claims/778758

MCP

get_claim(id=778758)