DeepSeek open-sources DSec: elastic sandbox infrastructure for agentic training

DeepSeek Elastic Compute (DSec)

DeepSeek published a paper on DSec, their internal sandbox system for training agents at scale. The idea is to let models practice with real tools in isolated environments that scale elastically. It handles 100K concurrent sandboxes, 11-second startup latency, and roughly $3 per sandbox. The post doesn't mention a code repo—only the arXiv paper is available so far.

Why it matters: DeepSeek open-sourced their internal agent-training sandbox infra with hard engineering numbers: 100k concurrent sandboxes, 11s cold start, $3/instance. Not a model release, so it stays below 85, but as a practical agent-training infra reference it's high-value for practitioners.