reasoning benchmark
AutomationBench Leaderboard
AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows.
Updated Oct 23, 2026
Models27
Model coverage27
MetricScore
EvidenceB
AutomationBench Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelDE | Score60.50% | Percentile100.00% | Participants27 | EvidenceC | Evaluated |
| Rank02 | ModelMA | Score59.90% | Percentile96.15% | Participants27 | EvidenceC | Evaluated |
| Rank03 | ModelDE | Score54.80% | Percentile92.31% | Participants27 | EvidenceC | Evaluated |
| Rank04 | ModelSA | Score53.80% | Percentile88.46% | Participants27 | EvidenceC | Evaluated |
| Rank05 | ModelIN | Score52.50% | Percentile84.62% | Participants27 | EvidenceC | Evaluated |
| Rank06 | ModelGO | Score51.30% | Percentile80.77% | Participants27 | EvidenceC | Evaluated |
| Rank07 | ModelME | Score49.40% | Percentile76.92% | Participants27 | EvidenceC | Evaluated |
| Rank08 | ModelZA | Score48.80% | Percentile73.08% | Participants27 | EvidenceC | Evaluated |
| Rank09 | ModelZA | Score48.20% | Percentile69.23% | Participants27 | EvidenceC | Evaluated |
| Rank10 | ModelAN | Score44.70% | Percentile65.38% | Participants27 | EvidenceC | Evaluated |
| Rank11 | ModelOP | Score41.40% | Percentile61.54% | Participants27 | EvidenceC | Evaluated |
| Rank12 | ModelAN | Score40.00% | Percentile57.69% | Participants27 | EvidenceC | Evaluated |
| Rank13 | ModelRA | Score39.30% | Percentile53.85% | Participants27 | EvidenceC | Evaluated |
| Rank14 | ModelTE | Score32.10% | Percentile50.00% | Participants27 | EvidenceC | Evaluated |
| Rank15 | ModelDE | Score31.80% | Percentile46.15% | Participants27 | EvidenceC | Evaluated |
| Rank16 | ModelAN | Score31.40% | Percentile42.31% | Participants27 | EvidenceC | Evaluated |
| Rank17 | ModelMA | Score30.80% | Percentile38.46% | Participants27 | EvidenceC | Evaluated |
| Rank18 | ModelGO | Score30.40% | Percentile34.62% | Participants27 | EvidenceC | Evaluated |
| Rank19 | ModelAC | Score27.30% | Percentile30.77% | Participants27 | EvidenceC | Evaluated |
| Rank20 | ModelAN | Score26.00% | Percentile26.92% | Participants27 | EvidenceC | Evaluated |
| Rank21 | ModelDE | Score25.70% | Percentile23.08% | Participants27 | EvidenceC | Evaluated |
| Rank22 | ModelDE | Score25.10% | Percentile19.23% | Participants27 | EvidenceC | Evaluated |
| Rank23 | ModelOP | Score18.10% | Percentile15.38% | Participants27 | EvidenceC | Evaluated |
| Rank24 | ModelAN | Score17.40% | Percentile11.54% | Participants27 | EvidenceC | Evaluated |
| Rank25 | ModelOP | Score15.20% | Percentile7.69% | Participants27 | EvidenceC | Evaluated |
| Rank26 | ModelOP | Score14.90% | Percentile3.85% | Participants27 | EvidenceC | Evaluated |
| Rank27 | ModelAN | Score13.50% | Percentile0.00% | Participants27 | EvidenceC | Evaluated |
AutomationBench Highlights
The leading models and scores on this benchmark.
AutomationBench Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for AutomationBench
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
What is AutomationBench?
What AutomationBench measures and how its scores work.
AutomationBench is a tool-use benchmark for evaluating AI agents on real-world workflow automation.
It measures agents' ability to orchestrate tools and complete multi-step automation tasks using the Score metric in ratio units.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- AutomationBench
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- overall
LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about AutomationBench.
Which model scores highest on AutomationBench?
DeepSeek-V4.1-Flash-1017 is currently ranked first with 60.50%.
What are the top three models on AutomationBench?
The current leaders are DeepSeek-V4.1-Flash-1017 (60.50%), Mistral Large 4 (59.90%), and DeepSeek-V4.1-Flash (54.80%).
Which AutomationBench model has the lowest official input price?
GLM-5.3-Flash has the lowest matched official input price at $0.12 input / $0.42 output per 1M tokens.
Which models are fastest among AutomationBench results?
The fastest matched records are Gemini 3.7 Flash (409.96 tok/s via Google), GPT-5.6 Luna (215.52 tok/s via OpenAI), and GLM-5.3-Flash (44.69 tok/s via FriendliAI).
Does the highest score result prove overall model quality?
No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.