FDE-Bench: leaderboard

Metric: Resolved rate (%; share of the 136 deployment-configuration tasks, 66 greenfield and 70 repair, whose redeployed artifacts pass all four gated check layers; one episode per task, same four-tool scaffold (bash, write_file, read_file, submit), task-specific step, time and token caps; submitted artifacts are rebuilt and redeployed in a pristine sandbox and graded by four gated programmatic check layers (build, readiness, behavior, spec conformance); resolved only if all four pass). Source: arxiv.org. Saturation forecast: Around December 2026. 7 models tracked.

Top models

#ModelScore
1Gemini 3.6 Flash75
2Claude Sonnet 5 (High)72.8
3DeepSeek V4 Pro69.9
4GPT-5.6 Terra69.9
5DeepSeek V4 Flash66.2
6GPT-5.6 Luna63.2
7Gemini 3.5 Flash Lite52.9

Interactive version: theaggregate.ai/benchmark?slug=fde-bench · How It Works · Data refreshed daily, snapshot 2026-09-26.