AgMoDB
ModelsAgentsOrchestratorsEvalsCompositesVisualizeIndustry
AgMoDB by @mistakeknot
All Benchmarks

Exploitbench

Exploitbench benchmark scores from Epoch AI's benchmark data collection.

aggregatePartial metadataepoch_ai

This definition has not completed editorial and provenance review. Verify its metric, version, source, and evaluation configuration upstream before using it for a model-selection decision.

Strengths

Sourced from Epoch AI's systematic benchmark data collection covering frontier models.

Caveats

Auto-generated definition — editorial description not yet written.

How to interpret scores

Contamination risk: moderateFreshness: periodic

Higher is better.

Caveat to keep in mind

Auto-generated definition — editorial description not yet written.

Leaderboard

Last synced: Nov 22, 2026, 8:17 AM
#ModelProviderScorePrice/M
1
Anthropic: Claude Haiku 4.5
source epoch_ai
Evidence observed Jul 25, 2026
Anthropic2.1$2.00
2
Claude Mythos Preview
source epoch_ai
Evidence observed Nov 22, 2026
Anthropic0.7—
3
Claude Opus 4.7
source epoch_ai
Evidence observed Nov 22, 2026
Anthropic0.2$10.00
4
Claude Sonnet 4.6 (Non-reasoning, High)
source epoch_ai
Evidence observed Nov 22, 2026
Anthropic0.2$6.00
5
Kimi K2.6 (Reasoning)
source epoch_ai
Evidence observed Nov 22, 2026
Kimi0.2$1.71
6
GLM-5.1 (Reasoning)
source epoch_ai
Evidence observed Nov 22, 2026
Z AI0.2$1.83
7
Gemini 3.1 Pro Preview
source epoch_ai
Evidence observed Nov 22, 2026
Google0.2$4.50
8
Claude 4.5 Haiku (Reasoning)
source epoch_ai
Evidence observed Nov 22, 2026
Anthropic0.1$2.00
9
MiniMax-M2.7
source epoch_ai
Evidence observed Nov 22, 2026
MiniMax0.1$0.53