DeepSeek has released DeepSeek-V4.1-Flash-1017, an updated AI model that the company says delivers stronger benchmark performance, faster task completion and shorter, more cost-effective outputs than its predecessor. [1, 5]
The model shows improvements in reasoning, coding and agent-related tasks, with notable gains on Terminal-Bench 4.0 and DeepSWE v1.1. [2, 4] DeepSeek describes the update as a “Pareto improvement” in the balance between performance and cost. [1, 5]
The release is seen as a move to sharpen competition with OpenAI, Anthropic and Google. [6] DeepSeek has also opened its Search and Fetch APIs to all developers. [1]