DeepSeek proposes Super-agent architecture: a two-tier design with Router and Worker
The official DeepSeek Harness team released "Super-agent Design Exploration," proposing a two-tier architecture composed of a Router layer and a Worker layer. The Router handles human interaction, maintains global context, and dispatches tasks, while Workers execute concurrently in isolated sessions. The team says the architecture has four characteristics—continuous, efficient, proactive, and persistent—and incorporates mechanisms such as event-driven operation, Inline & Spill, chain recall, speculative execution, and Watchdog. In preliminary experiments, the p50 and p95 task completion times for Super-agent were 12.4 minutes and 15.7 minutes, respectively, while the p95 for a single-session approach under neutral prompts was 29.5 minutes; the team notes that the experiments were based on synthetic workloads and a simplified environment. The article also provides a prototype implementation named dsh-superagent-prototype and makes clear that it is not an official DeepSeek product.
Why read this
AI score 70
DeepSeek has disclosed the two-tier architecture and preliminary experimental data of Super-agent, which can be understood against existing single-session approaches as a scheduling strategy for parallel multitasking.
Topics
Body · AI translation
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.
English text is being prepared.