Unverified50% confidenceBenchmarkExact time
DeepSeek V4.1 Flash scored 74 on the Deep SWE software engineering code test, making it the first open-weight model the UP host has seen to reach 74 on Deep SWE.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/16/2026
First Seen
Valid until: 12/15/2026
Sources
Related Entities
Related Claims
UnverifiedDeepSeek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77% similarPartially VerifiedDeepSeek V4.1 Flash scores lower than the previous version on GPQA, HLE, and other benchmarks.77% similarUnverifiedDeepSeek V4.1 Flash has topped the leaderboard on AA's new proprietary evaluation benchmark, surpassing Astra.75% similarUnverifiedDeepSeek V4.1 Flash was in closed beta during testing, but could be accessed normally by directly calling the model ID.75% similarUnverifiedDeepSeek V4.1 Flash adopts a Causal Encoder-Decoder asymmetric architecture with a total parameter count of 552B.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/924489API
curl https://kongchang.com/api/v1/knowledge/claims/924489MCP
get_claim(id=924489)