DeepSeek personally steps in to develop Claude Code's competitor, internally confirming the formation of a Harness team

robot
Abstract generation in progress

According to Beating monitoring, DeepSeek senior researcher Chen Deli confirmed on social media that the company is internally assembling a brand-new Harness team, directly benchmarking Anthropic’s Claude Code, with a focus on developing “DeepSeek Code Harness.”

The official recruitment page shows that DeepSeek currently has two key roles open in Haidian District, Beijing: Harness Product Manager and R&D Engineer. It also clearly states that new hires will take part in the entire development process of the “DeepSeek Desktop Agent Product.”

The job description defines the core path as “Model + Harness = Agent.” DeepSeek assigns engineering actions such as context management, tool calling, file reading and writing, terminal execution, and testing feedback—everything beyond the model itself—entirely to the Harness category.

Before this, the developer community had already independently rolled out an open-source terminal project aimed at benchmarking Claude Code: DeepSeek-TUI. With the company itself personally stepping in to build the Harness team, it means DeepSeek is reclaiming leadership over the terminal entry point, seeking to inject code testing feedback from real projects directly into the underlying model’s evolution closed loop.

View Original
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pinned