{"api_version":"1","generated_at":"2026-08-13T16:57:58+00:00","cve":"CVE-2026-73557","urls":{"html":"https://cve.report/CVE-2026-73557","api":"https://cve.report/api/cve/CVE-2026-73557.json","docs":"https://cve.report/api","cve_org":"https://www.cve.org/CVERecord?id=CVE-2026-73557","nvd":"https://nvd.nist.gov/vuln/detail/CVE-2026-73557"},"summary":{"title":"vLLM: Incomplete CVE-2025-62164 remediation can be bypassed by concurrent prompt parts","description":"vLLM is an inference and serving engine for large language models. From 0.20.2rc0 until 0.26.0, safe_load_prompt_embeds in vllm/renderers/embed_utils.py uses torch.sparse.check_sparse_tensor_invariants, whose process-global save, enable, and restore state can be raced by concurrent prompt_embeds parts submitted to POST /v1/chat/completions through AsyncMultiModalItemTracker.resolve_items, asyncio.gather, and the default executor, allowing an invalid sparse tensor to reach tensor.to_dense despite the CVE-2025-62164 guard when enable_prompt_embeds is enabled. This issue is fixed in version 0.26.0.","state":"PUBLISHED","assigner":"GitHub_M","published_at":"2026-08-13 15:20:18","updated_at":"2026-08-13 15:20:18"},"problem_types":["CWE-362","CWE-362 CWE-362: Concurrent Execution using Shared Resource with Improper Synchronization ('Race Condition')"],"metrics":[{"version":"4.0","source":"security-advisories@github.com","type":"Secondary","score":"6.3","severity":"MEDIUM","vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/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","data":{"version":"4.0","vectorString":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/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","baseScore":6.3,"baseSeverity":"MEDIUM","attackVector":"NETWORK","attackComplexity":"LOW","attackRequirements":"PRESENT","privilegesRequired":"NONE","userInteraction":"NONE","vulnConfidentialityImpact":"NONE","vulnIntegrityImpact":"NONE","vulnAvailabilityImpact":"LOW","subConfidentialityImpact":"NONE","subIntegrityImpact":"NONE","subAvailabilityImpact":"NONE","exploitMaturity":"NOT_DEFINED","confidentialityRequirement":"NOT_DEFINED","integrityRequirement":"NOT_DEFINED","availabilityRequirement":"NOT_DEFINED","modifiedAttackVector":"NOT_DEFINED","modifiedAttackComplexity":"NOT_DEFINED","modifiedAttackRequirements":"NOT_DEFINED","modifiedPrivilegesRequired":"NOT_DEFINED","modifiedUserInteraction":"NOT_DEFINED","modifiedVulnConfidentialityImpact":"NOT_DEFINED","modifiedVulnIntegrityImpact":"NOT_DEFINED","modifiedVulnAvailabilityImpact":"NOT_DEFINED","modifiedSubConfidentialityImpact":"NOT_DEFINED","modifiedSubIntegrityImpact":"NOT_DEFINED","modifiedSubAvailabilityImpact":"NOT_DEFINED","Safety":"NOT_DEFINED","Automatable":"NOT_DEFINED","Recovery":"NOT_DEFINED","valueDensity":"NOT_DEFINED","vulnerabilityResponseEffort":"NOT_DEFINED","providerUrgency":"NOT_DEFINED"}},{"version":"4.0","source":"CNA","type":"DECLARED","score":"6.3","severity":"MEDIUM","vector":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N","data":{"attackComplexity":"LOW","attackRequirements":"PRESENT","attackVector":"NETWORK","baseScore":6.3,"baseSeverity":"MEDIUM","privilegesRequired":"NONE","subAvailabilityImpact":"NONE","subConfidentialityImpact":"NONE","subIntegrityImpact":"NONE","userInteraction":"NONE","vectorString":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N","version":"4.0","vulnAvailabilityImpact":"LOW","vulnConfidentialityImpact":"NONE","vulnIntegrityImpact":"NONE"}}],"references":[{"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-pr7f-p5mw-fc87","name":"https://github.com/vllm-project/vllm/security/advisories/GHSA-pr7f-p5mw-fc87","refsource":"security-advisories@github.com","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://github.com/vllm-project/vllm/commit/793cf79c89d4049124e756915468ac30318f2e50","name":"https://github.com/vllm-project/vllm/commit/793cf79c89d4049124e756915468ac30318f2e50","refsource":"security-advisories@github.com","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://github.com/vllm-project/vllm/releases/tag/v0.26.0","name":"https://github.com/vllm-project/vllm/releases/tag/v0.26.0","refsource":"security-advisories@github.com","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://github.com/vllm-project/vllm/pull/48583","name":"https://github.com/vllm-project/vllm/pull/48583","refsource":"security-advisories@github.com","tags":[],"title":"","mime":"","httpstatus":"","archivestatus":"0"},{"url":"https://www.cve.org/CVERecord?id=CVE-2026-73557","name":"CVE Program record","refsource":"CVE.ORG","tags":["canonical"]},{"url":"https://nvd.nist.gov/vuln/detail/CVE-2026-73557","name":"NVD vulnerability detail","refsource":"NVD","tags":["canonical","analysis"]}],"affected":[{"source":"CNA","vendor":"vllm-project","product":"vllm","version":"affected >= 0.20.2rc0, < 0.26.0","platforms":[]}],"timeline":[],"solutions":[],"workarounds":[],"exploits":[],"credits":[],"nvd_cpes":[],"vendor_comments":[],"enrichments":{"kev":null,"epss":null,"legacy_qids":[]},"source_records":{"cve_program":{"containers":{"cna":{"affected":[{"product":"vllm","vendor":"vllm-project","versions":[{"status":"affected","version":">= 0.20.2rc0, < 0.26.0"}]}],"descriptions":[{"lang":"en","value":"vLLM is an inference and serving engine for large language models. From 0.20.2rc0 until 0.26.0, safe_load_prompt_embeds in vllm/renderers/embed_utils.py uses torch.sparse.check_sparse_tensor_invariants, whose process-global save, enable, and restore state can be raced by concurrent prompt_embeds parts submitted to POST /v1/chat/completions through AsyncMultiModalItemTracker.resolve_items, asyncio.gather, and the default executor, allowing an invalid sparse tensor to reach tensor.to_dense despite the CVE-2025-62164 guard when enable_prompt_embeds is enabled. This issue is fixed in version 0.26.0."}],"metrics":[{"cvssV4_0":{"attackComplexity":"LOW","attackRequirements":"PRESENT","attackVector":"NETWORK","baseScore":6.3,"baseSeverity":"MEDIUM","privilegesRequired":"NONE","subAvailabilityImpact":"NONE","subConfidentialityImpact":"NONE","subIntegrityImpact":"NONE","userInteraction":"NONE","vectorString":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N","version":"4.0","vulnAvailabilityImpact":"LOW","vulnConfidentialityImpact":"NONE","vulnIntegrityImpact":"NONE"}}],"problemTypes":[{"descriptions":[{"cweId":"CWE-362","description":"CWE-362: Concurrent Execution using Shared Resource with Improper Synchronization ('Race Condition')","lang":"en","type":"CWE"}]}],"providerMetadata":{"dateUpdated":"2026-08-13T15:00:15.746Z","orgId":"a0819718-46f1-4df5-94e2-005712e83aaa","shortName":"GitHub_M"},"references":[{"name":"https://github.com/vllm-project/vllm/security/advisories/GHSA-pr7f-p5mw-fc87","tags":["x_refsource_CONFIRM"],"url":"https://github.com/vllm-project/vllm/security/advisories/GHSA-pr7f-p5mw-fc87"},{"name":"https://github.com/vllm-project/vllm/pull/48583","tags":["x_refsource_MISC"],"url":"https://github.com/vllm-project/vllm/pull/48583"},{"name":"https://github.com/vllm-project/vllm/commit/793cf79c89d4049124e756915468ac30318f2e50","tags":["x_refsource_MISC"],"url":"https://github.com/vllm-project/vllm/commit/793cf79c89d4049124e756915468ac30318f2e50"},{"name":"https://github.com/vllm-project/vllm/releases/tag/v0.26.0","tags":["x_refsource_MISC"],"url":"https://github.com/vllm-project/vllm/releases/tag/v0.26.0"}],"source":{"advisory":"GHSA-pr7f-p5mw-fc87","discovery":"UNKNOWN"},"title":"vLLM: Incomplete CVE-2025-62164 remediation can be bypassed by concurrent prompt parts"}},"cveMetadata":{"assignerOrgId":"a0819718-46f1-4df5-94e2-005712e83aaa","assignerShortName":"GitHub_M","cveId":"CVE-2026-73557","datePublished":"2026-08-13T15:00:15.746Z","dateReserved":"2026-08-12T20:53:46.380Z","dateUpdated":"2026-08-13T15:00:15.746Z","state":"PUBLISHED"},"dataType":"CVE_RECORD","dataVersion":"5.2"},"nvd":{"publishedDate":"2026-08-13 15:20:18","lastModifiedDate":"2026-08-13 15:20:18","problem_types":["CWE-362","CWE-362 CWE-362: Concurrent Execution using Shared Resource with Improper Synchronization ('Race Condition')"],"metrics":{"cvssMetricV40":[{"source":"security-advisories@github.com","type":"Secondary","cvssData":{"version":"4.0","vectorString":"CVSS:4.0/AV:N/AC:L/AT:P/PR:N/UI:N/VC:N/VI:N/VA:L/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","baseScore":6.3,"baseSeverity":"MEDIUM","attackVector":"NETWORK","attackComplexity":"LOW","attackRequirements":"PRESENT","privilegesRequired":"NONE","userInteraction":"NONE","vulnConfidentialityImpact":"NONE","vulnIntegrityImpact":"NONE","vulnAvailabilityImpact":"LOW","subConfidentialityImpact":"NONE","subIntegrityImpact":"NONE","subAvailabilityImpact":"NONE","exploitMaturity":"NOT_DEFINED","confidentialityRequirement":"NOT_DEFINED","integrityRequirement":"NOT_DEFINED","availabilityRequirement":"NOT_DEFINED","modifiedAttackVector":"NOT_DEFINED","modifiedAttackComplexity":"NOT_DEFINED","modifiedAttackRequirements":"NOT_DEFINED","modifiedPrivilegesRequired":"NOT_DEFINED","modifiedUserInteraction":"NOT_DEFINED","modifiedVulnConfidentialityImpact":"NOT_DEFINED","modifiedVulnIntegrityImpact":"NOT_DEFINED","modifiedVulnAvailabilityImpact":"NOT_DEFINED","modifiedSubConfidentialityImpact":"NOT_DEFINED","modifiedSubIntegrityImpact":"NOT_DEFINED","modifiedSubAvailabilityImpact":"NOT_DEFINED","Safety":"NOT_DEFINED","Automatable":"NOT_DEFINED","Recovery":"NOT_DEFINED","valueDensity":"NOT_DEFINED","vulnerabilityResponseEffort":"NOT_DEFINED","providerUrgency":"NOT_DEFINED"}}]},"configurations":[]},"legacy_mitre":{"record":{"CveYear":"2026","CveId":"73557","Ordinal":"1","Title":"vLLM: Incomplete CVE-2025-62164 remediation can be bypassed by c","CVE":"CVE-2026-73557","Year":"2026"},"notes":[{"CveYear":"2026","CveId":"73557","Ordinal":"1","NoteData":"vLLM is an inference and serving engine for large language models. From 0.20.2rc0 until 0.26.0, safe_load_prompt_embeds in vllm/renderers/embed_utils.py uses torch.sparse.check_sparse_tensor_invariants, whose process-global save, enable, and restore state can be raced by concurrent prompt_embeds parts submitted to POST /v1/chat/completions through AsyncMultiModalItemTracker.resolve_items, asyncio.gather, and the default executor, allowing an invalid sparse tensor to reach tensor.to_dense despite the CVE-2025-62164 guard when enable_prompt_embeds is enabled. This issue is fixed in version 0.26.0.","Type":"Description","Title":"vLLM: Incomplete CVE-2025-62164 remediation can be bypassed by c"}]}}}