# CWE-Bench leaderboard

Agentic benchmark

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/leaderboards/benchmarks/agentic-cwe-bench

Every model the catalog carries a published CWE-Bench value for, ranked by that value.

## Benchmark record

| Field | Value |
| --- | --- |
| Category | Agentic |
| Measure | Pass@1 |
| Tasks | Real-world vulnerability patching tasks |
| Difficulty | Automated vulnerability remediation |
| Published by | [CWE-Bench leaderboard](https://cwe-bench.com/#leaderboard) |

An external benchmark for evaluating whether coding agents can produce correct patches for real-world software vulnerabilities.

## CWE-Bench ranking

1 model with a published CWE-Bench value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

| Rank | Model | Provider | Pass@1 | Evidence |
| --- | --- | --- | --- | --- |
| 1 | [Gemini 3.8 Flash Cyber](https://modelscale.dev/models/gemini-3-8-flash-cyber) | Google | 47.2 | Observed 2026-11-21 · source benchlm:models |

Rows are ordered by the value the source published, highest first. The source does not state whether a higher value is the better result, so this document does not either: for a benchmark that measures a rate of failure, read the table from the bottom.
