Benchmark Almanac

Last Secure Code Benchmark

Claude Fable 5 25.1% leads since Oct 2026 Not in model cards

Agents rebuild code that a real vulnerability once affected, from a specification that never mentions security. Hidden tests check function and replay an exploit.

450 tasks from 150 vulnerability scenarios, 115 from real CVEs and 35 expert written. Seven languages and 24 CWE classes. Each scenario appears at function, file and repository level. The headline metric counts a task only when both the function and security suites pass. Apache 2.0, also on Harbor Hub and Docker Hub.

In the authors’ words

AI coding agents now write much of the code that ships, yet whether that code is secure remains unclear. We introduce the Last Secure Code Benchmark, a comprehensive evaluation of secure code generation by coding agents. Built from 150 real vulnerabilities, its 450 tasks span seven widely used languages (Java, C, C++, JavaScript, Python, PHP, and Go), 24 CWEs including 21 of the 2025 CWE Top 25 and 8 of the Top 10 KEV weaknesses, more than any benchmark we compare with, and three vibe-coding scenarios: filling in a function, writing a file, and building a repository's source tree from scratch. Every task asks only for functionality, through a specification that never mentions security, and hidden functional and security tests grade the result. Across six frontier agents, 73.6% to 91.3% of tasks pass functionally, yet even the strongest, Claude Fable 5, produces code that is both functional and secure for only 25.1% of tasks, and the others for 16.4% to 24.7%. The best model only passes 12% of the vulnerability scenarios across all vibe coding setting, posing strong concern on the security of the code written by vibe coding.

Abstract of Last Secure Code Benchmark: Is the Code You Vibe Secure? (arXiv:2610.14766, CC0)
Made by Alibaba Qwen Team , Alibaba Cloud security and National University of Singapore Released Oct 2026 Secure code and review All tasks public Last model added Aug 2026 Scores refreshed 2 Nov 2026
Level
Expert Specialists with years of training, often over hours of work
Scored
Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
Answers
Code A program or patch run against tests.
alibaba/last-secure-code-benchmark 132 stars Paper Leaderboard last-secure-code-benchmark.github.io
6
Models scored on its own leaderboard

Claude Fable 5 leads with 25.1% among models released by Aug 2026

Best score per model from the benchmark's own leaderboard, read on 2 Nov 2026. Where a model ran with several settings or harnesses, we keep its best.

The newest model with a score here came out in Aug 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 25.1 Claude Fable 5
  2. 24.7 GPT-5.6 Sol
  3. 22.0 Qwen 3.8 Max
  4. 21.3 GLM-5.3
  5. 18.0 DeepSeek-V4-Pro
  6. 16.4 Kimi K3
  1. 1 Claude Fable 5 25.1%
  2. 2 GPT-5.6 Sol 24.7%
  3. 3 Qwen 3.8 Max 22%
  4. 4 GLM-5.3 21.3%
  5. 5 DeepSeek-V4-Pro 18%
  6. 6 Kimi K3 16.4%
#ModelScoreSettingDateSource
1 Claude Fable 5 25.1% Claude Code Jun 2026 Official leaderboard
2 GPT-5.6 Sol 24.7% Codex Jul 2026 Official leaderboard
3 Qwen 3.8 Max 22% Claude Code Aug 2026 Official leaderboard
4 GLM-5.3 21.3% Claude Code Aug 2026 Official leaderboard
5 DeepSeek-V4-Pro 18% Claude Code Apr 2026 Official leaderboard
6 Kimi K3 16.4% Claude Code Jul 2026 Official leaderboard

Number one over time

Lab claims count from the day the card came out, independent runs from the model’s release.

Anthropic
Own leaderboard 1 leader
Claude Fable 5 25.1%

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

On the benchmark’s own leaderboard Best so far
0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Fable 5 DeepSeek-V4-Pro 0% 20% 40% 60% 80% 100% 2026 Near the ceiling (85%) Claude Fable 5 DeepSeek-V4-Pro

No lab model card reports it yet

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.

Versions

Replaces SecCodeBench

More in Cybersecurity

All 111 in Cybersecurity
  • ExploitGym In use 869 real-vulnerability exploit-development tasks (userspace, V8, Linux kernel). Success means capturing a flag through the intended vulnerability. GPT-6 Astra 42.4%
  • ExploitBench In use 41 recent V8 JavaScript-engine N-day vulnerabilities. Progress measured by 16 grader-verified capability flags across five tiers. GPT-6 Astra 100%
  • CWE-bench In use Agents audit a repository and write a patch for a security flaw, without being told where the bug is. Covers six programming languages. Gemini 4 Carbon 71%
  • Real-world vulnerability discovery In use Agents search popular open-source codebases to rediscover about 1,200 recent confirmed vulnerabilities across many languages. Reports share discovered. Gemini 4 Argon 86%
  • Artificial Analysis Cyber Index In use Agents do enterprise cyber defense tasks from three benchmarks, CWE-Bench-AA, DeepsecBench-AA and CyberGym-E2E-AA, combined into one success rate. Mistral Large 4 (Le Chonk, open weights) 51%
  • Big Sleep Evaluation (Google) In use Agents search large real codebases such as Chrome and Safari for critical, hard to find vulnerabilities. Scored by pass@1 success rate, run without safety guardrails. Gemini 3.5 Flash Cyber 72%