vLLM CVE-2026-54233
MEDIUM
GHSA-6pr9-rp53-2pmc
Severity by source
AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
vLLM enforces no native authentication; advisory explicitly states unauthenticated access, warranting PR:N over the official PR:L; impact is pure availability with no confidentiality or integrity effect.
Primary rating from Vendor (https://github.com/vllm-project/vllm).
CVSS VectorVendor: https://github.com/vllm-project/vllm
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H
Lifecycle Timeline
2Blast Radius
ecosystem impact- 4 pypi packages depend on vllm (3 direct, 1 indirect)
Ecosystem-wide dependent count for version 0.23.0.
DescriptionCVE.org
Summary
vLLM's /v1/audio/transcriptions endpoint limits compressed upload size but not decoded PCM output. A 25MB OPUS file expands to ~14.9GB of float32 PCM at decode time. Tested on vLLM v0.19.0.
Details
SpeechToTextProcessor rejects uploads over VLLM_MAX_AUDIO_CLIP_FILESIZE_MB (default 25MB) based on compressed byte length, but the audio decoder in audio.py accumulates all decoded frames into memory with no size limit before returning:
# speech_to_text.py L184-189
if len(audio_data) / 1024 ** 2 > self.max_audio_filesize_mb:
raise VLLMValidationError(...)
y, sr = load_audio(buf, sr=self.asr_config.sample_rate)
# decoded size unchecked
# audio.py L77-107
chunks: list[npt.NDArray] = []
for frame in container.decode(stream):
chunks.append(frame.to_ndarray())
audio = np.concatenate(chunks, axis=-1).astype(np.float32)
# single contiguous allocationA 25MB OPUS file at 6kbps encodes ~8.7 hours of audio. Decoding produces ~5.7GB of float32 PCM (232x amplification), and np.concatenate then allocates a second contiguous array, bringing peak RSS to ~14.9GB from a single request. SpeechToTextConfig.max_audio_clip_s (default 30s) applies only after the full decode and does not prevent the allocation.
Impact
An unauthenticated attacker can exhaust server memory with a small number of concurrent requests, each a valid upload within the documented size limit. Severity was assessed with reference to prior OOM vulnerability reports in vLLM.
Fix
A fix for this vulnerability was merged here: https://github.com/vllm-project/vllm/pull/44970
AnalysisAI
Memory exhaustion via audio decompression bomb in vLLM's /v1/audio/transcriptions endpoint allows a remote attacker to crash or severely degrade the inference server by uploading a crafted OPUS file that passes the 25MB compressed-size check but decodes to ~14.9GB of float32 PCM in memory. Affected are vLLM installations through v0.23.0 with speech-to-text (ASR) functionality enabled. …
Unlock full vulnerability intelligence
- Risk assessment & exploitation conditions
- Attack chain visualization
- Remediation with exact patch versions
- Threat intelligence from 22 sources
- Personal watchlist & email alerts
Free forever · No credit card required
Attack ChainAIDerived
Hypothetical attack flow derived from CVE metadata
Vulnerability AssessmentAI
| Exploitation | The `/v1/audio/transcriptions` endpoint must be enabled, which requires vLLM to be configured with an ASR model - deployments serving only LLM inference are not affected. … Additional conditions and limiting factors are described in the full assessment. |
| Risk Assessment | The CVSS 3.1 base score of 6.5 (AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H) reflects a network-exploitable, low-complexity availability-only impact. … Full risk analysis with EPSS, KEV, and SSVC signal comparison available after sign-in. |
| Exploit Scenario | An attacker with network access to the vLLM API endpoint creates a valid OPUS audio file of ~25MB encoded at ~6kbps, embedding approximately 8.7 hours of audio content that comfortably fits within vLLM's default compressed-size limit. The attacker submits this file via a standard multipart POST to `/v1/audio/transcriptions`; the compressed-size check passes, and the server begins decoding - allocating ~5.7GB for decoded PCM frame chunks and an additional ~9.2GB for the `np.concatenate()` contiguous array. … |
| Remediation | Upgrade to vLLM v0.23.1rc0 or a subsequent GA release that incorporates PR #44970 (https://github.com/vllm-project/vllm/pull/44970, commit 1b1359c33269446f13c05da9a90c25174cbea590). … Detailed patch versions, workarounds, and compensating controls in full report. |
Threat intelligence, references, and detailed analysis are available after sign-in.
Wazuh SIEM platform versions 4.4.0 through 4.9.0 contain an unsafe deserialization vulnerability in the DistributedAPI t
BentoML version 1.4.2 and earlier contains an unauthenticated remote code execution vulnerability through insecure deser
pgAdmin 4 contains critical remote code execution vulnerabilities in the Query Tool download and Cloud Deployment endpoi
The renderLocalView function in render/views.py in graphite-web in Graphite 0.9.5 through 0.9.10 uses the pickle Python
BentoML is a Python library for building online serving systems optimized for AI apps and model inference. Rated critica
OpenSSL before 0.9.8za, 1.0.0 before 1.0.0m, and 1.0.1 before 1.0.1h does not properly restrict processing of ChangeCiph
pyLoad download manager version prior to 0.5.0b3.dev77 exposes the Flask SECRET_KEY through an unauthenticated endpoint.
In Mercurial before 4.1.3, "hg serve --stdio" allows remote authenticated users to launch the Python debugger, and conse
Unauthenticated remote code execution in Marimo ≤0.20.4 allows attackers to execute arbitrary system commands via the `/
pyLoad is the free and open-source Download Manager written in pure Python. Rated medium severity (CVSS 5.3), this vulne
Langflow (a visual LLM pipeline builder) contains a critical unauthenticated code execution vulnerability (CVE-2026-3301
Cross-user flow execution in Langflow (< 1.9.1) lets any authenticated API-key holder run another user's flow by passing
Same technique Information Disclosure
View allShare
External POC / Exploit Code
Leaving vuln.today