DeepSeek Launches Updated AI Model With Faster Performance and Shorter OutputsDeepSeek Launches Updated AI Model With Faster Performance and Shorter Outputs

Chinese artificial intelligence company DeepSeek has released an updated AI model that it says delivers faster performance and shorter outputs, intensifying competition with US rivals Anthropic and OpenAI as well as domestic developers, including Moonshot AI and Z.ai, Saritha Rai reported for Bloomberg News.

DeepSeek introduces its updated V4.1-Flash-1017 AI model as competition in the global artificial intelligence market intensifies. [Photo: Tim Reckmann Flickr]

The company introduced DeepSeek-V4.1-Flash-1017 over the weekend, announcing the update through a series of posts on X.

DeepSeek said the new version offers stronger benchmark performance, faster task completion and shorter responses than its predecessor.

The updated model is available through DeepSeek’s website, mobile application and application programming interface (API) for developers, according to the company’s official announcement.

DeepSeek said the latest version delivers improvements in reasoning, coding, agent-based tasks and visual benchmarks.

Its release reflects the rapid pace of development among Chinese AI companies seeking to compete with established US providers.

The competition also highlights differences in the business models used by AI developers. Many Chinese companies promote models with publicly available weights, allowing developers to download and adapt them under the applicable license terms.

Several leading US companies, meanwhile, rely on proprietary models accessed through commercial products and APIs.

DeepSeek’s latest update adds to a series of releases aimed at improving model capabilities and efficiency.

Its performance claims, however, should be evaluated through independent testing across comparable tasks, hardware configurations and costs before conclusions can be drawn about how it compares with competing systems.