SShogo Bench

ExploitBench

Exploit development across tiers of difficulty.

Agent Index weight
0%
shown for context only
Models tested
9
from the benchmark’s leaderboard and reports
Normalised for the index
0 → 0, 1 → 100
Official site

Results

#ModelScoreNormalisedHarnessCost/taskDate
1Claude Mythos Preview
Anthropic
73.8%source——
2GPT-5.5 Supported
OpenAI
47.4%source——
3Claude Opus 4.7 Supported
Anthropic
26.5%source——
4Gemini 3.1 Pro Preview Supported
Google
26.1%source——
5Claude Sonnet 4.6 Supported
Anthropic
23.6%source——
6Kimi K2.6 Estimated
Moonshot AI
18.4%source——
7GLM-5.1 Estimated
Z.ai (Zhipu)
18.1%source——
8Claude Haiku 4.5 Estimated
Anthropic
13.7%source——
9MiniMax-M2.7
MiniMax
13.3%source——