CVE-2025-67729: high-severity vulnerability in InternLM lmdeploy
lmdeploy vulnerable to Arbitrary Code Execution via Insecure Deserialization in torch.load()
Published
21Vexday Risk Score
No sign of exploitation. No public exploitation artifact known so far.
ssvc Trackcvss 8.8epss 0.6%
exploitation probability
0.6%top 53% of all CVEs
observed exploitation
nono source reports it
LMDeploy is a toolkit for compressing, deploying, and serving LLMs. Prior to version 0.11.1, an insecure deserialization vulnerability exists in lmdeploy where torch.load() is called without the weights_only=True parameter when loading model checkpoint files. This allows an attacker to execute arbitrary code on the victim's machine when they load a malicious .bin or .pt model file. This issue has been patched in version 0.11.1.
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:H/A:H
Affected products
InternLM · lmdeployRelated CVEs — InternLM lmdeploy
In the same product, most dangerous first.
CVE-2026-33626HIGHLMDeploy Vulnerable to Server-Side Request Forgery (SSRF) via Vision-Language Image LoadingEPSS 1.5%CVE-2026-76850CRITICALLMDeploy Remote Code Execution via Unsafe Pickle Deserialization in the Disaggregated Serving Peer ConnectorEPSS 1.3%CVE-2025-59953CRITICALLMdeploy has Remote Code Execution by Pickle Deserialization via zmq_rpc.call_and_response() in InterLM/lmdeployEPSS 0.8%CVE-2026-92971HIGHInternLM LMDeploy through 0.17.0 Assertion Denial of ServiceEPSS 0.7%CVE-2025-66455CRITICALLMDeploy has Remote Code Execution by Pickle Deserialization via handle_zmq_recv in lmdeploy/lmdeploy/pytorch/disagg/conn/engine_conn.pyEPSS 0.7%CVE-2026-92983HIGHInternLM LMDeploy through 0.17.0 Memory Exhaustion via Session ID MismatchEPSS 0.7%