FDE-Bench - Orchestration: leaderboard
The 48 orchestration tasks of FDE-Bench (Ding et al., 2026), which ask an agent to author or fix Compose stacks and Kubernetes manifests so that services connect and become ready. Agents share one four-tool scaffold; submitted Dockerfiles, manifests and monitoring configuration are rebuilt and redeployed in a clean sandbox and pass only if the build, readiness, behavior and specification-conformance check layers all pass, with no LLM judge. Resolved rate in percent; higher is better.
Metric: Resolved rate (%; the 48 orchestration-axis tasks (Compose stacks and Kubernetes manifests); 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 February 2027. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3.6 Flash | 72.9 | #72 |
| 2 | DeepSeek V4 Pro | 68.8 | #181 |
| 3 | Claude Sonnet 5 (High) | 68.8 | #108 (Claude Sonnet 5) |
| 4 | GPT-5.6 Terra | 64.6 | #60 |
| 5 | Gemini 3.5 Flash Lite | 62.5 | #300 |
| 6 | GPT-5.6 Luna | 60.4 | #117 |
| 7 | DeepSeek V4 Flash | 58.3 | #254 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=fde-bench-orchestration · How It Works · Data refreshed daily, snapshot 2026-11-12.