Deepseek V4.1 Flash 1017 智能体任务超越更大模型并大幅降低成本
原标题：Deepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costs
AI 摘要
中国 AI 公司 Deepseek 发布 Deepseek-V4.1-Flash-1017，通过网站、应用和 API 提供，并在 Hugging Face 与 ModelScope 开源模型权重。该模型采用带自反馈的强化学习（RLSF）进行后训练，运行更快、输出更短，在内部基准的编程与智能体任务上超过 Claude Opus 5、GPT 5.6-Sol 等更大模型，且单任务成本显著更低。Deepseek 同时首次向开发者开放其内部搜索引擎，推出 Search 与 Fetch API。
正文节选
Deepseek V4.1 Flash 1017 beats larger rivals in agentic tasks while once again massively cutting per-task costs Key Points - Chinese AI company Deepseek has released Deepseek-V4.1-Flash-1017 on its platforms and as an open-weight model on Hugging Face and ModelScope. - Trained with reinforcement learning using self-feedback, the model runs faster, produces shorter outputs, and performs well in end-to-end software development and agent tasks, the company says. - In internal benchmarks, the update