AutomationBench

Tests whether the AI can run real multi-step business workflows — the kind of end-to-end office processes companies want to automate — from start to finish. Higher is better.

DeepSeek-V4.1-Flash-1017 by DeepSeek holds the best AutomationBench result among tracked models, at 60.5% (released Oct 17 2026). Higher scores are better on this benchmark.

1. DeepSeek-V4.1-Flash-1017 (DeepSeek) — 60.5%; 2. Mistral Large 4 (Mistral) — 59.9%; 3. DeepSeek-V4.1-Flash (DeepSeek) — 54.8%; 4. Gemini 4 Argon (Google) — 51.3%; 5. Qwen3.8-Max-0902 (Qwen) — 50.8%.

19 tracked models have a published AutomationBench score. Scores are collected from published lab reports and benchmark sources, with recorded sources credited alongside the results. We do not re-run the evaluations.

DeepSeek-V4.1-Flash-1017 by DeepSeek is the highest-ranked model with downloadable weights on AutomationBench, scoring 60.5% at rank 1 overall.