DeepSeek Releases V4.1-Flash-1017, Claiming Coding Performance Edge at Fraction of Competitor Cost

DeepSeek has released DeepSeek-V4.1-Flash-1017, an updated open-weight AI model that the company claims outperforms larger rivals on coding benchmarks while cutting per-task costs to under $0.20. Released on October 17, 2026, the model scores 77.0% on the DeepSWE v1.1 software engineering benchmark, ahead of Claude Opus 5 (74.0%) and GPT-5.6 Sol (73.0%) on the same measure. The model is now available via DeepSeek Chat, the company's API, and as open-source weights under an MIT license on Hugging Face and ModelScope.