Benchmark

AutomationBench

Published by Zapier over its own automation surface, and cards report different task subsets of it.

Released
2026-03-10
Categories
agent
Openness
unknown
Source
model_reports
Reported scores
5

Reported scores

model_reports

ModelOrganizationReported valueReportedEvidence
Claude Opus 5Anthropic26.02026-07-24evidence
DeepSeek-V4-Pro-0813DeepSeek31.82026-08-13evidence
GLM-5.3-FlashZ.ai48.82026-08-27evidence
Hy4 previewTencent32.12026-08-28evidence
Kimi K3Moonshot AI30.82026-06-13evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.