ONNX: Heap-Buffer-Overflow READ in Gemm Version Converter Adapter via Undersized Input Shape
CVE-2026-63632 · GHSA-p893-rvq9-2xf9 · published 2026-07-24
- Severity
- Low
- CVSS score
- 3.3
- Chance of exploitation (EPSS)
- 0.17%
- Fixed in
- 1.22.0
- Exploitation Public proof of concept
- Automatable No
- Technical impact Partial
- CISA SSVC
Any application that uses onnx.versionconverter.convertversion() on untrusted models is vulnerable. This includes model conversion pipelines and tools that auto-downgrade opset versions for compatibility. On Release builds the OOB read is silent — the read value propagates into the converted model's output shape, potentially leaking heap data. On ASan builds it's detected as a heap-buffer-overflow. Could also cause crashes with different heap layouts.
Upgrade onnx to 1.22.0 or later.
| Package | Registry | Affected | Fixed in |
|---|---|---|---|
| onnx | PyPI | >= 1.3.0, <= 1.21.0 | 1.22.0 |
- CriticalLiteLLM has SQL Injection in Proxy API key verificationlitellmExploited
- Criticaln8n Vulnerable to Remote Code Execution via Expression Injectionn8n, n8nExploited
- HighLiteLLM has a sandbox escape in custom-code guardraillitellmExploited
- HighLangflow affected by Remote Code Execution via validate_code() exec()langflowExploited
- CriticalJupyter MCP Server: Browser DNS rebinding can hijack the Jupyter backend through management routesjupyter-mcp-server
- Criticalmem0ai Databricks vector store vulnerable to SQL injectionmem0ai
- HighONNX: Malicious ONNX models can crash servers by exploiting unprotected object settings.onnx
- HighONNX: TOCTOU arbitrary file read/write in save_external_datonnx
- Highonnx Vulnerable to Path Traversal via Symlinkonnx
- HighONNX Untrusted Model Repository Warnings Suppressed by silent=True in onnx.hub.load() — Silent Supply-Chain Attackonnx
- HighOpen Neural Network Exchange (ONNX) Path Traversal Vulnerabilityonnx
- Highonnx allows Arbitrary File Overwrite in download_model_with_test_dataonnx
The impact, patch and workaround text comes from the advisory, shortened (GitHub Advisory Database, CC BY 4.0). Proof-of-concept and reproduction sections and code are left out; the original advisory has them.