CVE-2026-47749

HIGHPre-NVD 7.87.8
EchelonGraph scoreMEDIUM confidence

This high-severity CVE scores 7.8 under a secondary CVSS source (NVD's own analysis pending). EPSS exploit probability: 0.2%, top 90% of all CVEs by exploit prediction. GitHub Security Advisory data not yet ingested — confidence will rise once GHSA publishes (typical lag: hours to days for open-source ecosystem CVEs; never for infrastructure-only CVEs).

Triggered by: NVD CVSS baseline
Sources: epss, secondary
7.8
EchelonGraph verdictPlan a fixSerious severity, but no confirmed exploitation yet.
  • High severity, but no confirmed exploitation yet
CISA-KEV: Not listedEPSS: 0%CVSS: 7.8Exploit: NoneExposed: 0

No vendor fix yet — apply a workaround or compensating control (WAF / firewall / segmentation) and watch for a patch.

stable-diffusion.cpp is a pure C/C++ library for running diffusion model (Stable Diffusion, Flux, Wan, Qwen Image, Z-Image, and more) inference. Versions prior to master-584-0a7ae07 are vulnerable to heap buffer overflow in SHORT_BINUNICODE parsing for PyTorch checkpoint files. The pickle .ckpt parser in src/model.cpp contained a heap buffer overflow vulnerability in the SHORT_BINUNICODE opcode handler. The issue was caused by sign confusion on the opcode length field. A crafted .ckpt file could trigger memcpy with a very large length derived from a negative signed value, causing immediate heap corruption. Any application using affected stable-diffusion.cpp releases to load untrusted .ckpt model files could be vulnerable. A malicious checkpoint file could cause heap corruption through memcpy with an attacker-controlled length. This may lead to process crash and could potentially be leveraged for code execution depending on heap layout. The attack requires the victim or application to load a .ckpt file from an untrusted source, such as a downloaded model from a model sharing site. The issue has been resolved in version master-584-0a7ae07. If developers are unable to immediately update their applications they can work around this issue by not loading .ckpt checkpoint files from untrusted sources, and referring to trusted model sources and safer formats such as .safetensors where possible.

CVSS v3
7.8
EG Score
7.8(medium)
EPSS
10.3%
KEV
Not listed

Published

June 16, 2026

Last Modified

June 25, 2026

Advisory Details (3)

Auto-updated Jun 16, 2026
🔬 Proof of concept available. Patch available. Sources: github_commit, github_pr, github.
github Patch Available🟡 PoC Available

Heap buffer overflow in SHORT_BINUNICODE parsing for PyTorch checkpoint files · Advisory · leejet/stable-diffusion.cpp · GitHub

https://github.com/leejet/stable-diffusion.cpp/security/advisories/GHSA-2c29-5hxg-fv9g
github_pr

feat: add restricted torch legacy checkpoint loading

Fix merged in leejet/stable-diffusion.cpp PR #1443 on 2026-04-19 — awaiting tagged release

https://github.com/leejet/stable-diffusion.cpp/pull/1443
github_commit

commit 0a7ae07f948e (leejet/stable-diffusion.cpp)

Fix landed in leejet/stable-diffusion.cpp commit 0a7ae07f948e — awaiting tagged release

https://github.com/leejet/stable-diffusion.cpp/commit/0a7ae07f948eff4611968a65a22bd7c7031ad74f

Weakness Classification(2)

MITRE Common Weakness Enumeration — the root-cause categories this CVE belongs to.

Data Freshness Timeline

(refreshed 20× in last 7d / 85× in last 30d)

Each row is a source pipeline that fetched or updated this CVE on that date, with what changed. For example, "NVD update" means NVD published or revised its analysis for this CVE; "MITRE cvelistV5" means we ingested or refreshed it from the CNA feed. Most recent first.

  1. 2026-07-13 23:22 UTCEG score recompute
  2. 2026-07-13 22:30 UTCEPSS rescore
  3. 2026-07-13 11:05 UTCEG score recompute
  4. 2026-07-13 06:13 UTCEPSS rescore
  5. 2026-07-12 22:00 UTCEG score recompute
  6. 2026-07-12 09:18 UTCEG score recompute
  7. 2026-07-12 05:46 UTCEPSS rescore
  8. 2026-07-11 21:01 UTCEG score recompute
  9. 2026-07-11 08:45 UTCEG score recompute
  10. 2026-07-11 08:27 UTCEPSS rescore
  11. 2026-07-10 20:29 UTCEG score recompute
  12. 2026-07-10 08:13 UTCEG score recompute
  13. 2026-07-09 19:44 UTCEG score recompute
  14. 2026-07-09 19:10 UTCEPSS rescore
  15. 2026-07-09 01:59 UTCEG score recompute
  16. 2026-07-08 15:16 UTCEPSS rescore
  17. 2026-07-08 13:20 UTCEG score recompute
  18. 2026-07-08 00:00 UTCEG score recompute
  19. 2026-07-07 13:46 UTCEPSS rescore
  20. 2026-07-07 11:42 UTCEG score recompute
  21. 2026-07-06 23:08 UTCEG score recompute
  22. 2026-07-06 16:27 UTCEPSS rescore
  23. 2026-07-06 02:23 UTCEPSS rescore
  24. 2026-07-05 22:33 UTCEG score recompute
  25. 2026-07-05 10:17 UTCEG score recompute
Show 60 more
  1. 2026-07-05 02:30 UTCEPSS rescore
  2. 2026-07-04 21:27 UTCEG score recompute
  3. 2026-07-04 09:08 UTCEG score recompute
  4. 2026-07-04 06:31 UTCEPSS rescore
  5. 2026-07-03 19:21 UTCEG score recompute
  6. 2026-07-03 07:06 UTCEG score recompute
  7. 2026-07-02 18:46 UTCEG score recompute
  8. 2026-07-02 06:30 UTCEG score recompute
  9. 2026-07-01 18:13 UTCEG score recompute
  10. 2026-07-01 15:07 UTCEPSS rescore
  11. 2026-07-01 05:56 UTCEG score recompute
  12. 2026-06-30 23:22 UTCEPSS rescore
  13. 2026-06-30 17:27 UTCEG score recompute
  14. 2026-06-30 05:02 UTCEG score recompute
  15. 2026-06-29 16:35 UTCEG score recompute
  16. 2026-06-29 14:06 UTCEPSS rescore
  17. 2026-06-29 14:06 UTCEPSS rescore
  18. 2026-06-29 04:19 UTCEG score recompute
  19. 2026-06-28 15:25 UTCEG score recompute
  20. 2026-06-28 14:07 UTCEPSS rescore
  21. 2026-06-28 04:56 UTCEPSS rescore
  22. 2026-06-28 04:56 UTCEPSS rescore
  23. 2026-06-28 03:09 UTCEG score recompute
  24. 2026-06-27 14:53 UTCEG score recompute
  25. 2026-06-27 03:08 UTCEPSS rescore
  26. 2026-06-27 02:37 UTCEG score recompute
  27. 2026-06-26 14:19 UTCEG score recompute
  28. 2026-06-26 02:03 UTCEG score recompute
  29. 2026-06-25 13:49 UTCEPSS rescore
  30. 2026-06-25 13:46 UTCEG score recompute
  31. 2026-06-25 01:29 UTCEG score recompute
  32. 2026-06-24 14:05 UTCEPSS rescore
  33. 2026-06-24 13:13 UTCEG score recompute
  34. 2026-06-24 00:44 UTCEG score recompute
  35. 2026-06-23 12:14 UTCEG score recompute
  36. 2026-06-22 23:57 UTCEG score recompute
  37. 2026-06-22 14:25 UTCEPSS rescore
  38. 2026-06-22 14:25 UTCEPSS rescore
  39. 2026-06-22 11:29 UTCEG score recompute
  40. 2026-06-21 23:00 UTCEG score recompute
  41. 2026-06-21 14:56 UTCEPSS rescore
  42. 2026-06-21 14:56 UTCEPSS rescore
  43. 2026-06-21 10:34 UTCEG score recompute
  44. 2026-06-21 01:59 UTCEPSS rescore
  45. 2026-06-21 01:59 UTCEPSS rescore
  46. 2026-06-20 22:18 UTCEG score recompute
  47. 2026-06-20 09:55 UTCEG score recompute
  48. 2026-06-19 21:35 UTCEG score recompute
  49. 2026-06-19 19:26 UTCEPSS rescore
  50. 2026-06-19 19:26 UTCEPSS rescore
  51. 2026-06-19 09:17 UTCEG score recompute
  52. 2026-06-18 21:00 UTCEG score recompute
  53. 2026-06-18 17:52 UTCEPSS rescore
  54. 2026-06-18 17:52 UTCEPSS rescore
  55. 2026-06-18 08:41 UTCEG score recompute
  56. 2026-06-17 20:23 UTCEG score recompute
  57. 2026-06-17 17:53 UTCEPSS rescore
  58. 2026-06-17 08:07 UTCEG score recompute
  59. 2026-06-16 19:50 UTCEG score recompute
  60. 2026-06-16 19:48 UTCNVD updatefirst tracked

Frequently asked(5)

What is CVE-2026-47749?
CVE-2026-47749 is a high vulnerability published on June 16, 2026. stable-diffusion.cpp is a pure C/C++ library for running diffusion model (Stable Diffusion, Flux, Wan, Qwen Image, Z-Image, and more) inference. Versions prior to master-584-0a7ae07 are vulnerable to heap buffer overflow in SHORTBINUNICODE parsing for PyTorch checkpoint files. The pickle .ckpt…
When was CVE-2026-47749 disclosed?
CVE-2026-47749 was first published in the National Vulnerability Database on June 16, 2026, with the most recent update on June 25, 2026. EchelonGraph re-ingests CVE updates from NVD on a 2-hour cycle, so this page reflects the latest published state.
Is CVE-2026-47749 actively exploited?
CVE-2026-47749 is not currently on CISA's Known Exploited Vulnerabilities catalog. FIRST EPSS estimates a 10.3% percentile likelihood of exploitation in the next 30 days — higher percentiles indicate greater predicted risk.
What is the CVSS score of CVE-2026-47749?
CVE-2026-47749 has a CVSS v3 base score of 7.8 (NVD).
How do I remediate CVE-2026-47749?
Patch to the fixed version published by the affected vendor. Where vendor advisories exist for CVE-2026-47749, EchelonGraph cross-links them in the Vendor Advisories panel below — those typically contain the canonical remediation steps, fixed version numbers, and any vendor-specific mitigations.

Dependency Blast Radius

Explore the affected products and dependency analysis for CVE-2026-47749

Explore →

Is Your Infrastructure Affected by CVE-2026-47749?

EchelonGraph automatically scans your cloud infrastructure and maps CVE exposure using blast radius analysis.