CVE-2026-93840 json
vLLM before 0.29.0 Cross-Request Logits Corruption via allowed_token_ids
published 2026-09-18 51 d ago modified 2026-09-29 41 d ago also PYSEC-2026-3998 , EUVD-2026-83284
- cvss
- 3.7
- version 3.1, CNA
- cisa kev
- Not listed
- not in CISA's catalog
- epss
- 0.2%
- EPSS 0.2% chance of exploitation in 30 days, percentile 15%
- public exploits
- 0
- none known
Summary
vLLM before 0.29.0 validates allowed_token_ids against tokenizer length instead of model output logits width in SamplingParams._validate_allowed_token_ids(). Attackers can supply token IDs above the output vocabulary that pass validation, causing LogitBiasState to corrupt GPU logits state and allow concurrent requests to sample tokens outside their allowlists.
Patch / fix
- pypi vllm 0.29.0
- vllm-project vllm 0.29.0
- github.com/vllm-project/vllm@5b0e5b6 commit
- github.com/vllm-project/vllm#49080 pull request
Vendor advisories
What the vendors themselves say about their products: affected, fixed or not affected, and what to do.
red hat product security vex CVE-2026-93840 updated cvss 4.8
vllm: vLLM: Information disclosure via incorrect token ID validation
-
Red Hat AI Inference Server rhaii/vllm-cpu-rhel9, rhaii/vllm-tpu-rhel9, rhaii/vllm-cuda-rhel9, rhaii/vllm-rocm-rhel9, rhaiis/vllm-cpu-rhel9, rhaiis/vllm-tpu-rhel9, +7 more Red Hat Enterprise Linux AI (RHEL AI) 3 rhelai3/bootc-cuda-rhel9, rhelai3/bootc-rocm-rhel9, rhelai3/bootc-gaudi-rhel9, rhelai3/bootc-aws-cuda-rhel9, rhelai3/bootc-gcp-cuda-rhel9, rhelai3/bootc-azure-cuda-rhel9, +1 more Red Hat OpenShift AI (RHOAI) rhoai/odh-vllm-gaudi-rhel9, rhoai/odh-kserve-agent-rhel9, rhoai/odh-kserve-router-rhel9, rhoai/odh-llm-d-kv-cache-rhel9, rhoai/odh-kserve-controller-rhel9, rhoai/odh-th-torch-cuda-py312-rhel9, +1 moreaffected: no fix available “Fix deferred”
Red Hat CSAF VEX © Red Hat, Inc., from https://security.access.redhat.com/data/csaf/v2/vex/; changed: reduced to product statuses and remediations. CC BY 4.0 · csaf
CISA SSVC
- Exploitation
- none
- Automatable
- no
- Technical impact
- partial
- Assessed
- 48 d ago
Exploit prediction
0.25% chance of exploitation in the next 30 days, higher than 15% of all scored CVEs (FIRST EPSS, 2026-11-08).
CVSS scores
| assigned by | version | score | vector |
|---|---|---|---|
| CNA | 4.0 | 6.3 | CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N |
| CNA | 3.1 | 3.7 | CVSS:3.1/AV:N/AC:H/PR:N/UI:N/S:U/C:N/I:L/A:N |
| CNA | 4.0 | 6.3 | CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:L/VA:N/SC:N/SI:N/SA:N/E:X/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X |
| NVD | 3.1 | 5.3 | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N |
| OSV | 3.1 | 5.3 | CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:L/A:N |
Affected packages
| ecosystem | package | affected | fixed in |
|---|---|---|---|
| pypi | vllm | <0.29.0 |
0.29.0 |
Affected products
- vllm-project vllm
Classification
- Weakness
-
- CWE-129 Improper Validation of Array Index
References
-
github.com
/vllm-project/vllm/pull/49080
patch, issue-tracking
-
github.com
/vllm-project/vllm/commit/5b0e5b69ac1a3884a6479c9537789c95263cc804
patch
-
github.com
/vllm-project/vllm
product, package
-
github.com
/vllm-project/vllm/blob/v0.28.0/vllm/sampling_params.py#L881-L903
technical-description, product, web
-
github.com
/vllm-project/vllm/blob/v0.28.0/vllm/v1/worker/gpu/sample/logit_bias.py#L179-L191
technical-description, product, web
-
vulncheck.com
/advisories/vllm-before-0.29.0-cross-request-logits-corruption-via-allowed-token-ids
third-party-advisory, advisory
Sources
- CVE List CVE-2026-93840
- NVD CVE-2026-93840
- OSV PYSEC-2026-3998
Fencer Vulns keeps one entry per vulnerability: the CVE id wins, then GHSA, and records that alias each other are merged, except that two CVEs always keep separate entries. The same data is available as JSON.