{"id":"CVE-2026-54233","aliases":["EUVD-2026-38404","GHSA-6pr9-rp53-2pmc","PYSEC-2026-3404"],"status":"published","title":"vLLM: OOM Denial of Service via Audio Decompression Bomb","description":"vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, 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. This vulnerability is fixed in 0.23.1rc0.","severity":{"version":"3.1","score":6.5,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","level":"medium","assessor":"cna","from":"cve"},"epss":{"day":"2026-11-10","score":0.00422,"percentile":0.36815,"model":"v2026.06.15"},"facts":{"in_kev":false,"known_exploited":false,"public_exploit":false,"fix_available":true},"cwes":["CWE-409"],"published":"2026-06-17T14:06:22Z","modified":"2026-10-28T02:30:53Z","fixes":[{"ecosystem":"PyPI","package":"vllm","fixed_in":["0.24.0"],"sources":["ghsa","osv"]}],"affected":[{"source":"ghsa","ecosystem":"PyPI","package":"vllm","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"last_affected":"0.23.0"},{"fixed":"0.24.0"}]}]},{"source":"osv","ecosystem":"PyPI","package":"vllm","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"fixed":"0.24.0"}]}],"versions":["0.0.1","0.1.0","0.1.1","0.1.2","0.1.3","0.1.4","0.1.5","0.1.6","0.1.7","0.10.0","0.10.1","0.10.1.1","0.10.2","0.11.0","0.11.1","0.11.2","0.12.0","0.13.0","0.14.0","0.14.1","0.15.0","0.15.1","0.16.0","0.17.0","0.17.1","0.18.0","0.18.1","0.19.0","0.19.1","0.2.0","0.2.1","0.2.1.post1","0.2.2","0.2.3","0.2.4","0.2.5","0.2.6","0.2.7","0.20.0","0.20.1","0.20.2","0.21.0","0.22.0","0.22.1","0.23.0","0.3.0","0.3.1","0.3.2","0.3.3","0.4.0","0.4.0.post1","0.4.1","0.4.2","0.4.3","0.5.0","0.5.0.post1","0.5.1","0.5.2","0.5.3","0.5.3.post1","0.5.4","0.5.5","0.6.0","0.6.1","0.6.1.post1","0.6.1.post2","0.6.2","0.6.3","0.6.3.post1","0.6.4","0.6.4.post1","0.6.5","0.6.6","0.6.6.post1","0.7.0","0.7.1","0.7.2","0.7.3","0.8.0","0.8.1","0.8.2","0.8.3","0.8.4","0.8.5","0.8.5.post1","0.9.0","0.9.0.1","0.9.1","0.9.2"]},{"source":"osv","ecosystem":"PyPI","package":"vllm","ranges":[{"type":"ECOSYSTEM","events":[{"introduced":"0"},{"last_affected":"0.23.0"}]}],"versions":["0.0.1","0.1.0","0.1.1","0.1.2","0.1.3","0.1.4","0.1.5","0.1.6","0.1.7","0.10.0","0.10.1","0.10.1.1","0.10.2","0.11.0","0.11.1","0.11.2","0.12.0","0.13.0","0.14.0","0.14.1","0.15.0","0.15.1","0.16.0","0.17.0","0.17.1","0.18.0","0.18.1","0.19.0","0.19.1","0.2.0","0.2.1","0.2.1.post1","0.2.2","0.2.3","0.2.4","0.2.5","0.2.6","0.2.7","0.20.0","0.20.1","0.20.2","0.21.0","0.22.0","0.22.1","0.23.0","0.3.0","0.3.1","0.3.2","0.3.3","0.4.0","0.4.0.post1","0.4.1","0.4.2","0.4.3","0.5.0","0.5.0.post1","0.5.1","0.5.2","0.5.3","0.5.3.post1","0.5.4","0.5.5","0.6.0","0.6.1","0.6.1.post1","0.6.1.post2","0.6.2","0.6.3","0.6.3.post1","0.6.4","0.6.4.post1","0.6.5","0.6.6","0.6.6.post1","0.7.0","0.7.1","0.7.2","0.7.3","0.8.0","0.8.1","0.8.2","0.8.3","0.8.4","0.8.5","0.8.5.post1","0.9.0","0.9.0.1","0.9.1","0.9.2"]}],"products":[{"vendor":"vllm","product":"vllm"},{"vendor":"vllm-project","product":"vllm"}],"references":[{"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-6pr9-rp53-2pmc","tags":["x_refsource_CONFIRM","Third Party Advisory","advisory","web"]},{"url":"https://github.com/vllm-project/vllm/pull/44970","tags":["x_refsource_MISC","Issue Tracking","Patch","fix","web"]},{"url":"https://github.com/vllm-project/vllm/commit/1b1359c33269446f13c05da9a90c25174cbea590","tags":["web"]},{"url":"https://github.com/vllm-project/vllm/releases/tag/v0.23.1rc0","tags":["web"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-54233","tags":["advisory"]},{"url":"https://github.com/advisories/GHSA-6pr9-rp53-2pmc","tags":["advisory"]},{"url":"https://github.com/pypa/advisory-database/tree/main/vulns/vllm/PYSEC-2026-3404.yaml","tags":["web"]},{"url":"https://github.com/vllm-project/vllm","tags":["package"]},{"url":"https://pypi.org/project/vllm","tags":["web","package"]},{"url":"https://github.com/CVEProject/cvelistV5/tree/main/cves/2026/54xxx/CVE-2026-54233.json","tags":["advisory"]}],"sources":{"cve":[{"id":"CVE-2026-54233","title":"vLLM: OOM Denial of Service via Audio Decompression Bomb","description":"vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, 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. This vulnerability is fixed in 0.23.1rc0.","status":"published","published":"2026-06-22T22:10:45.689Z","modified":"2026-06-23T12:15:43.607Z","cvss":[{"version":"3.1","score":6.5,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","level":"medium","assessor":"cna"}],"cwes":["CWE-409"],"updated_at":"2026-11-02T13:38:09.774038Z"}],"euvd":[{"id":"EUVD-2026-38404","description":"vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, 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. This vulnerability is fixed in 0.23.1rc0.","status":"published","published":"2026-06-22T22:10:45Z","modified":"2026-06-23T12:15:43Z","cvss":[{"version":"3.1","score":6.5,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","level":"medium"}],"updated_at":"2026-11-02T18:20:33.850201Z"}],"ghsa":[{"id":"GHSA-6pr9-rp53-2pmc","title":"vLLM: OOM Denial of Service via Audio Decompression Bomb","description":"### Summary\nvLLM'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.\n\n### Details\n`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:\n\n```python\n# speech_to_text.py L184-189\nif len(audio_data) / 1024 ** 2 \u003e self.max_audio_filesize_mb:\n    raise VLLMValidationError(...)\ny, sr = load_audio(buf, sr=self.asr_config.sample_rate)  # decoded size unchecked\n\n# audio.py L77-107\nchunks: list[npt.NDArray] = []\nfor frame in container.decode(stream):\n    chunks.append(frame.to_ndarray())\naudio = np.concatenate(chunks, axis=-1).astype(np.float32)  # single contiguous allocation\n```\n\nA 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.\n\n### Impact\nAn 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.\n\n### Fix\n\nA fix for this vulnerability was merged here: https://github.com/vllm-project/vllm/pull/44970","status":"published","published":"2026-06-17T14:06:22Z","modified":"2026-07-17T16:39:17Z","cvss":[{"version":"3.1","score":6.5,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","level":"medium"}],"cwes":["CWE-409"],"updated_at":"2026-11-02T14:40:44.915484Z"}],"nvd":[{"id":"CVE-2026-54233","description":"vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, 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. This vulnerability is fixed in 0.23.1rc0.","status":"published","published":"2026-06-22T23:16:31.007Z","modified":"2026-10-25T23:10:00.167Z","cvss":[{"version":"3.1","score":6.5,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","level":"medium","assessor":"secondary"}],"cwes":["CWE-409"],"updated_at":"2026-11-02T14:12:25.288279Z"}],"osv":[{"id":"CVE-2026-54233","title":"vLLM: OOM Denial of Service via Audio Decompression Bomb","description":"vLLM is an inference and serving engine for large language models (LLMs). Prior to 0.23.1rc0, 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. This vulnerability is fixed in 0.23.1rc0.","status":"published","published":"2026-06-22T22:10:45.689Z","modified":"2026-10-28T02:30:53.556964047Z","cvss":[{"version":"3.1","score":6.5,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","level":"medium"}],"cwes":["CWE-409"],"updated_at":"2026-11-02T15:45:10.818486Z"},{"id":"GHSA-6pr9-rp53-2pmc","title":"vLLM: OOM Denial of Service via Audio Decompression Bomb","description":"### Summary\nvLLM'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.\n\n### Details\n`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:\n\n```python\n# speech_to_text.py L184-189\nif len(audio_data) / 1024 ** 2 \u003e self.max_audio_filesize_mb:\n    raise VLLMValidationError(...)\ny, sr = load_audio(buf, sr=self.asr_config.sample_rate)  # decoded size unchecked\n\n# audio.py L77-107\nchunks: list[npt.NDArray] = []\nfor frame in container.decode(stream):\n    chunks.append(frame.to_ndarray())\naudio = np.concatenate(chunks, axis=-1).astype(np.float32)  # single contiguous allocation\n```\n\nA 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.\n\n### Impact\nAn 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.\n\n### Fix\n\nA fix for this vulnerability was merged here: https://github.com/vllm-project/vllm/pull/44970","status":"published","published":"2026-06-17T14:06:22Z","modified":"2026-09-10T03:51:08.015014698Z","cvss":[{"version":"3.1","score":6.5,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","level":"medium"}],"cwes":["CWE-409"],"updated_at":"2026-11-02T15:46:39.965326Z"},{"id":"PYSEC-2026-3404","title":"vLLM: OOM Denial of Service via Audio Decompression Bomb","description":"### Summary\nvLLM'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.\n\n### Details\n`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:\n\n```python\n# speech_to_text.py L184-189\nif len(audio_data) / 1024 ** 2 \u003e self.max_audio_filesize_mb:\n    raise VLLMValidationError(...)\ny, sr = load_audio(buf, sr=self.asr_config.sample_rate)  # decoded size unchecked\n\n# audio.py L77-107\nchunks: list[npt.NDArray] = []\nfor frame in container.decode(stream):\n    chunks.append(frame.to_ndarray())\naudio = np.concatenate(chunks, axis=-1).astype(np.float32)  # single contiguous allocation\n```\n\nA 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.\n\n### Impact\nAn 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.\n\n### Fix\n\nA fix for this vulnerability was merged here: https://github.com/vllm-project/vllm/pull/44970","status":"published","published":"2026-07-13T15:46:18.784318Z","modified":"2026-07-13T16:33:32.280562891Z","cvss":[{"version":"3.1","score":6.5,"vector":"CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H","level":"medium"}],"updated_at":"2026-11-02T15:52:56.847083Z"}]}}