Chinese AI startup DeepSeek is assembling a new team to develop Code Harness, a tool for autonomous programming. This was announced by the company’s engineer, Deli Chen.
🚀 We’re hiring! DeepSeek is forming a new Harness team to build Code Harness from the ground up—may be you can call it DeepSeek Code or something like this hhh🤣🤣🤣
📍 Based in Beijing. Two roles open:
🧠 Harness Product Manager → https://t.co/vb3aWbYV9L
👨💻 Harness R&D…— Deli Chen (@victor207755822) May 20, 2026
The developer posted two job openings on social media: a product manager and a research and development engineer. Both positions are based in Beijing.
The new solution will directly compete with Claude Code from Anthropic and Codex from OpenAI.
According to Chen, the team will build Code Harness “from scratch.” He described the project as “DeepSeek Code or something like that.”
The job descriptions use the formula “model + harness = agent.”
By harness, they mean a layer on top of the LLM. It is responsible for managing context, planning steps, memory, tool invocation, file handling, terminal operations, testing, and feedback.
DeepSeek aims to offer not just a code generation assistant, but an agent product for executing multi-step tasks in a development environment.
Candidates are expected to have experience with Claude Code, Codex, Cursor, GitHub Copilot, Manus, and OpenClaw. Requirements also include understanding agent cycles, MCP, multi-agent systems, context management, and vibe coding.
The team will work closely with DeepSeek researchers.
DeepSeek V4 supports integration with Claude Code through an Anthropic-compatible API. The V4 Flash version, introduced on April 24, costs $0.14 per 1 million input tokens. In comparison, Claude Opus 4.7 is priced at $15 for the same volume.
Back in February, Anthropic accused three Chinese AI startups—DeepSeek, Moonshot, and MiniMax—of a large-scale campaign to use Claude to enhance their own models.