Back to feed
MarkTechPost
MarkTechPost
7/27/2026
Kimi AI and kvcache-ai Open Sources ‘AgentENV’: A Distributed System that Powers Agentic Reinforcement Learning (RL) Training for Kimi K3

Kimi AI and kvcache-ai Open Sources ‘AgentENV’: A Distributed System that Powers Agentic Reinforcement Learning (RL) Training for Kimi K3

Short summary

Moonshot AI's Kimi team and kvcache-ai have open-sourced AgentENV (AENV) under the MIT license as part of Kimi K3 Open Day. AgentENV is a distributed system that runs agent sandboxes as Firecracker microVMs with millisecond snapshot, resume, and 16-way fork capabilities behind an E2B-compatible API. It is designed to power agentic reinforcement learning training for Kimi K3.

  • AgentENV open-sourced under MIT by Kimi team and kvcache-ai
  • Runs agent sandboxes as Firecracker microVMs with ms-level snapshot, resume, and 16-way fork
  • Exposes an E2B-compatible API for agentic RL training of Kimi K3

Generated with AI, which can make mistakes.

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more