DeepSeek Unveils DSec Sandbox Infrastructure for Large-Scale Agent Training
- Published
- Source
- TechNode.com
DeepSeek has introduced the DSec platform, a sandbox infrastructure tailored for extensive agent training. This innovative system integrates various sandbox types, including function-call, container, microVM, and full-VM sandboxes, while effectively managing their lifecycle alongside reinforcement learning tasks.
The significance of DSec lies in its ability to support large-scale operations, with a production-scale unit comprising approximately 160 nodes. It can handle around 3 million sandboxes daily and accommodate over 380,000 concurrent sandboxes, showcasing its robust capabilities in the field of agent training.
Looking ahead, the DSec platform's ability to create more than 5,000 sandboxes per second positions it as a critical tool for developers and researchers in the AI domain. No further timeline was disclosed at the time of publication.
Original report
DeepSeek details DSec sandbox infrastructure for agent training
TechNode.com · TechNode FeedThis briefing is an independently written summary based on publicly available reporting and is provided for industry information and news discovery. The original report and source publication are credited and linked where applicable. RobotToday does not claim ownership of third-party source material.
Rights concerns? Contact [email protected] with the relevant URL and details. We will review the matter and take appropriate action where warranted.
More related news
-
1
RobotToday Weekly November 9–13, 2026 RobotToday · 10 hours ago
- 2
- 3
-
4
RobotToday Weekly November 2 – 6, 2026 RobotToday · 1 week ago
- 5
- 6
- 7
- 8
- 9
- 10
- 11
- 12
Related Suppliers
Fireworks AI
Redwood City AI infra startup (2022) serving open-source LLMs like Kimi K3 via API, used by Cursor, Vercel and Notion.
Alpha Robotics 苏州阿尔法机器人有限公司
Suzhou, China commercial service robot maker founded 2006 with a Tokyo office; ships delivery, AMR and disinfection robots to 8,000 customers in 75 countries.
AMESA
Walnut Creek, California AI company founded 2023 as Composabl; sells a platform for building and deploying industrial autonomous agents.
Simate
Physical AI startup founded in 2026 whose Simate-beta robot model topped the RoboDojo leaderboard; also builds the Sinfra training platform.
sees.ai
UK drone autonomy firm; first to hold routine BVLOS CAA authorisation; OS for autonomous grid inspection; Boeing-backed.
e-Novia S.p.A.
Milan-based enterprise factory founded in 2015, incubating robotics and AI ventures including YAPE, Italy's first autonomous last-mile delivery robot.
FastLabel
Tokyo AI data-infrastructure company building robotics/physical-AI datasets and VLA-model pipelines, backed by Panasonic, Marubeni and NTT DOCOMO.
Metalware
San Francisco firmware security startup (YC W24) offering automated binary fuzzing to detect vulnerabilities in embedded device firmware.
Learning Robots
Learning Robots is a Paris-Saclay CNRS spin-off making AlphAI, an educational robot and software for teaching artificial intelligence.
BridgeDP Robotics 桥介数物
Shenzhen developer of universal robot 'cerebellum' motion-control algorithms for humanoid, quadruped and wheeled-legged robots using RL and motion capture.
Micromagic Systems Robotics Limited
UK robotics engineering firm founded 1999 by Matt Denton, specializing in animatronic, hydraulic, and custom robotic control systems for film, TV, and R&D.
MMC UAV (MicroMultiCopter Aero Technology) 深圳市多翼创新科技有限公司
Chinese industrial UAV maker founded 2009 in Shenzhen, offering drones for inspection, delivery, geospatial mapping, public safety, and drone shows globally.
Robotics needs a service framework.
RSF defines a common language for robot service capability, lifecycle operations, certification pathways, and service-provider networks.
Join the RobotToday community on LinkedIn
Daily robotics news, in-depth analysis, conference highlights, and discussions with professionals worldwide.