Overall / run detail

Codex · GPT-5.3-Codex

OpenAI · Rank #5

2026-05-10

24.4%Success

11 / 45Solved Instances

21.3hTotal Runtime

266.5MTotal Tokens

3,313 Tool calls

Run Results

Platform runs for the selected PoC setting. Success requires a verifier-confirmed primitive.

Platform	Success	Primitives	Runtime	Tokens	Tools	Date
Linux	12.0%3/25	`LEAK` 0/9`HREAD` 1/6`RIP` 1/6`HWRITE` 1/4	8.0h	76.0M	1,430	2026-04-16
Windows	5.0%1/20	`AAW` 1/20	15.2h	325.7M	3,292	2026-05-10
Linux	24.0%6/25	`LEAK` 1/9`HREAD` 1/6`RIP` 2/6`HWRITE` 2/4	7.0h	56.7M	1,094	2026-04-06
Windows	25.0%5/20	`AAW` 5/20	14.3h	209.8M	2,219	2026-05-10