CVE-2026-55646 — Uncontrolled Resource Consumption

Vendor	Product	Version range	Fixed in
vllm-project	vllm	—	—
vllm	vllm	>= 0.22.0 < 0.24.0	0.24.0
vllm	vllm	>= 0.22.0 < 0.24.0	0.24.0

Red Hat

vllm: vLLM: Denial of Service due to excessive memory allocation via oversized audio file uploads

vendor_redhat·2026-07-06·CVSS 6.5

CVE-2026-55646 [MEDIUM] CWE-770 vllm: vLLM: Denial of Service due to excessive memory allocation via oversized audio file uploads vllm: vLLM: Denial of Service due to excessive memory allocation via oversized audio file uploads vLLM is an inference and serving engine for large language models. From 0.22.0 to 0.23.0, the /v1/audio/transcriptions and /v1/audio/translations routes call request.file.read() to fully materialize an uploaded audio file into memory before vLLM checks the documented VLLM_MAX_AUDIO_CLIP_FILESIZE_MB compressed upload size limit (default 25 MB) later in the speech-to-text preprocessing step, so an API caller who can reach those routes can submit an oversized multipart upload and cause vLLM to allocate memory proportional to the uploaded file size before the request is rejected as too large, creating memory pressure or terminating the process depending on deployment resource limits. This issue i

GHSA

vLLM: Speech-to-text upload size limit is enforced after full UploadFile read

ghsa·2026-07-17

CVE-2026-55646 [MEDIUM] CWE-400 vLLM: Speech-to-text upload size limit is enforced after full UploadFile read vLLM: Speech-to-text upload size limit is enforced after full UploadFile read ## Summary Current-head vLLM documents `VLLM_MAX_AUDIO_CLIP_FILESIZE_MB` as the maximum audio file size accepted by the speech-to-text APIs. The default is 25 MB. `vllm/envs.py` also describes files larger than this value as rejected. The `/v1/audio/transcriptions` and `/v1/audio/translations` routes call `await request.file.read()` before vLLM checks that limit. In FastAPI and Starlette, `UploadFile.read()` returns bytes from the uploaded file object; when called without a size argument, the route materializes the remaining file contents. vLLM then performs the compressed file-size check later in `_preprocess_speech_to_text()` against the already-created `bytes` object. As a result, the documented compressed

VulDB

vllm-project vLLM up to 0.23.x Audio Transcription/Translation /v1/audio/transcriptions request.file.read File unrestricted upload

vuldb·2026-07-06·CVSS 6.5

CVE-2026-55646 [MEDIUM] vllm-project vLLM up to 0.23.x Audio Transcription/Translation /v1/audio/transcriptions request.file.read File unrestricted upload A vulnerability was found in vllm-project vLLM up to 0.23.x. It has been declared as critical. Affected by this vulnerability is the function request.file.read of the file /v1/audio/transcriptions of the component Audio Transcription/Translation Handler. Such manipulation of the argument File leads to unrestricted upload. This vulnerability is listed as CVE-2026-55646. The attack may be performed from remote. There is no available exploit.

CVE-2026-55646: vLLM is an inference and serving engine for large language models. From 0.22.0 to 0.23.0, the /v1/audio/transcriptions and /v1/audio/translations routes call…

Affected

CVSS provenance