AutomationBench
Multi-step computer automation and workflow-completion benchmark for agentic UI tasks.
Models ranked
13
tracked on this benchmark
Score band
60.5 – 12.9
best → lowest tracked
Snapshot trend
+14.90
Oct 16 → Oct 26 · 1 models
Leaderboard
Tracked models ranked by Accuracy (higher is better).
Notes: Self-reported on the Nanbeige4.3-33B model card; highest thinking effort, preserve_thinking=true. 600 publicly released tasks; no harness named. Retrieved 2026-10-26. Harness not stated on the model card.
Configuration: Gemini 4 Argon; private split
Notes: Blog table + methodology page: private set figure taken from Zapier's official AutomationBench leaderboard (Zapier-run). Note: most other catalog automation-bench rows are the public split; not directly comparable. Retrieved 2026-10-17.
Notes: Reflection first-party final model card (HF https://huggingface.co/reflection/Beam-1.0-501B-A23B, BF16 column), retrieved 2026-10-16. Replaces 2026-10-05 preview-blog score 37.0. Card label 'AutomationBench - public split only'. Lab self-reported.
Notes: Model variant: Kimi K3 (max reasoning effort). Benchmark variant: AutomationBench 600-task public subset. Harness/evaluator: official AutomationBench GitHub setup; evaluator not disclosed by Moonshot. Source posture: Moonshot/Kimi self-reported launch table; confidence: medium because the value is vendor-reported and not independently reproduced. Recommended seed value: 30.8.
Notes: Unstarred in Anthropic's shared Fable 5/Mythos 5 launch table, so this is a Claude Fable 5 score.
How to read this benchmark
This benchmark scores models where higher is better. Use scores for directional filtering and shortlisting, not universal quality ranking; then validate pricing, context window, provider availability, and fit for your workload.
Trust this score when
- There is a fresh timestamped snapshot (or multiple snapshots) for this benchmark.
- The model list covers the same version family you can actually deploy today.
- Top candidates overlap with your required routing and feature requirements.
Be cautious when
- There is only one benchmark snapshot or the dataset appears stale.
- The benchmark metric direction is opposite of your decision objective.
- The score difference between options is narrow and likely within implementation variance.
Related benchmarks
Last reviewed: Jun 15, 2026