AutomationBench-AA
Agentic SaaS workflow benchmark: automating multi-step business workflows across software tools.
Actively evaluated by Artificial Analysis on current frontier models; no saturation evidence recorded.
Performance Timeline
Longitudinal progression of model scores against human baselines.Performance & Historical Trajectory
Empirical score progression across model release dates and evaluation rounds.
| Model | Score | Date | Source Type | Provenance |
|---|---|---|---|---|
| Celeris-1 | 0.4% | 2026-11-15 | independent | Source ↗ |
| Ling 3.1 Flash | 61.7% | 2026-11-15 | independent | Source ↗ |
| NVIDIA Nemotron 3 Nano 4B | 0.4% | 2026-11-15 | independent | Source ↗ |
| K2 Horizon 0.9B | 0.2% | 2026-11-15 | independent | Source ↗ |
| Qwen3.8-Flash-Next | 55.9% | 2026-11-15 | independent | Source ↗ |
| Ministral 3 8B | 0.5% | 2026-11-15 | independent | Source ↗ |
| Llama 4 Maverick | 0.2% | 2026-11-15 | independent | Source ↗ |
| Qwen3.8 2.4T A95B | 57.2% | 2026-11-15 | independent | Source ↗ |
| Kimi K2.7 Code | 24.5% | 2026-11-15 | independent | Source ↗ |
| Gemini 4 Hydrogen | 76.1% | 2026-11-15 | independent | Source ↗ |
| DeepSeek V4-Pro | 56.7% | 2026-11-15 | independent | Source ↗ |
| A.X-K2 | 2.5% | 2026-11-15 | independent | Source ↗ |
| Qwen3.7 Plus | 17.4% | 2026-11-15 | independent | Source ↗ |
| Ling 3.0 Tiny | 0.3% | 2026-11-15 | independent | Source ↗ |
| Nemotron 3.5 Bolt 74B-A9B | 15.5% | 2026-11-15 | independent | Source ↗ |
| Muse Ember | 68.1% | 2026-11-15 | independent | Source ↗ |
| DeepSeek V4 Flash Vision | 47.5% | 2026-11-15 | independent | Source ↗ |
| Mercury 2.5 | 2.4% | 2026-11-15 | independent | Source ↗ |
| Gemini 3.8 Flash | 59.9% | 2026-11-15 | independent | Source ↗ |
| Inkling Small | 4.8% | 2026-11-15 | independent | Source ↗ |
| Ministral 3 14B | 0.6% | 2026-11-15 | independent | Source ↗ |
| Magistral Small 1.2 | 0.2% | 2026-11-15 | independent | Source ↗ |
| Gemma 4 26B A4B | 1.9% | 2026-11-15 | independent | Source ↗ |
| Ling-3.0-flash-Fin | 7.4% | 2026-11-15 | independent | Source ↗ |
| GLM 5.3 | 62.2% | 2026-11-15 | independent | Source ↗ |
| Qwen3 Coder Next | 1.1% | 2026-11-15 | independent | Source ↗ |
| Quasar 438B | 23.9% | 2026-11-15 | independent | Source ↗ |
| Ring-2.6-1T | 0.7% | 2026-11-15 | independent | Source ↗ |
| Muse Wonder Full | 71.3% | 2026-11-15 | independent | Source ↗ |
| Inkling 2 Small | 45.6% | 2026-11-15 | independent | Source ↗ |
| Nemotron 3.5 Super VL | 14.8% | 2026-11-15 | independent | Source ↗ |
| Mistral Medium 3.5 | 6.3% | 2026-11-15 | independent | Source ↗ |
| Agnes 2.5 Pro Theta | 48.8% | 2026-11-15 | independent | Source ↗ |
| MiniMax M3.1 Flash | 72.5% | 2026-11-15 | independent | Source ↗ |
| Hy4 | 65.6% | 2026-11-15 | independent | Source ↗ |
| Gemma 4 31B | 4.9% | 2026-11-15 | independent | Source ↗ |
| Beam 1.0 | 48.5% | 2026-11-15 | independent | Source ↗ |
| gpt-oss-120b | 0.2% | 2026-11-15 | independent | Source ↗ |
| GPT-5.5 Instant | 36.4% | 2026-11-15 | independent | Source ↗ |
| Gemini 3.1 Pro | 35.4% | 2026-11-15 | independent | Source ↗ |
| K2 Horizon 375B A23B | 37.2% | 2026-11-15 | independent | Source ↗ |
| Trinity Large Thinking | 1.3% | 2026-11-15 | independent | Source ↗ |
| JT-4.1 Flash 236B A21B | 36.7% | 2026-11-15 | independent | Source ↗ |
| Hy3 | 18.4% | 2026-11-15 | independent | Source ↗ |
| Qwen3.8 Max (0902) | 56.2% | 2026-11-15 | independent | Source ↗ |
| Granite 4.2 8B | 2.1% | 2026-11-15 | independent | Source ↗ |
| Llama 4 Scout | 0.2% | 2026-11-15 | independent | Source ↗ |
| K2 Horizon 7B | 20.9% | 2026-11-15 | independent | Source ↗ |
| MiMo-V2.6-Pro | 58.6% | 2026-11-15 | independent | Source ↗ |
| Grok 4.7 | 65.6% | 2026-11-15 | independent | Source ↗ |
| Claude Haiku 5.5 | 35.4% | 2026-11-15 | independent | Source ↗ |
| GPT-6 Astra | 68.5% | 2026-11-15 | independent | Source ↗ |
| North Mini Code | 1.5% | 2026-11-15 | independent | Source ↗ |
| GPT-6 Luna | 53.2% | 2026-11-15 | independent | Source ↗ |
| GLM 5.3 Flash | 60.4% | 2026-11-15 | independent | Source ↗ |
| K-EXAONE 2.0 | 3.9% | 2026-11-15 | independent | Source ↗ |
| Solar Pro 4 | 13.6% | 2026-11-15 | independent | Source ↗ |
| Mistral Small 4 | 1% | 2026-11-15 | independent | Source ↗ |
| MiMo-V2.6-Flash | 64.1% | 2026-11-15 | independent | Source ↗ |
| MiniMax M3 | 21.3% | 2026-11-15 | independent | Source ↗ |
| Granite 4.2 30B | 1.5% | 2026-11-15 | independent | Source ↗ |
| Qwen3.5 397B A17B | 6.9% | 2026-11-15 | independent | Source ↗ |
| Inkling | 5% | 2026-11-15 | independent | Source ↗ |
| Gemini 4 Carbon | 77.8% | 2026-11-15 | independent | Source ↗ |
| GPT-6.1 Terra | 60.2% | 2026-11-15 | independent | Source ↗ |
| Claude Fable 5.2 | 76.4% | 2026-11-15 | independent | Source ↗ |
| Granite 4.2 3B | 0.4% | 2026-11-15 | independent | Source ↗ |
| Step 5 | 51% | 2026-11-15 | independent | Source ↗ |
| Gemini 3.5 Flash-Lite | 25% | 2026-11-15 | independent | Source ↗ |
| GPT-6.1 Sol | 64.9% | 2026-11-15 | independent | Source ↗ |
| Apodex 1.1 | 28.2% | 2026-11-15 | independent | Source ↗ |
| Gemini 4 Argon | 77.5% | 2026-11-15 | independent | Source ↗ |
| Ministral 3 3B | 0.7% | 2026-11-15 | independent | Source ↗ |
| Kimi K3.1 Pro | 73.3% | 2026-11-15 | independent | Source ↗ |
| MiniCPM5-2B | 1.8% | 2026-11-15 | independent | Source ↗ |
| Solar Mini 4 | 22.3% | 2026-11-15 | independent | Source ↗ |
| LongCat 2.0 | 4.5% | 2026-11-15 | independent | Source ↗ |
| Nemotron 3 Ultra | 3% | 2026-11-15 | independent | Source ↗ |
| Muse Glimmer | 6.8% | 2026-11-15 | independent | Source ↗ |
| DeepSeek V4.1 Flash 1017 | 66.6% | 2026-11-15 | independent | Source ↗ |
| K2 Horizon MoVA 36B A4B | 26.1% | 2026-11-15 | independent | Source ↗ |
| Ling 3.1 Flash 1019 | 67.5% | 2026-11-15 | independent | Source ↗ |
| Nanbeige4.3-33B | 53.1% | 2026-11-15 | independent | Source ↗ |
| Claude Sonnet 5.5 | 71.8% | 2026-11-15 | independent | Source ↗ |
| Ling-3.0-flash-VL | 15.7% | 2026-11-15 | independent | Source ↗ |
| Nex-N2-Pro | 17.6% | 2026-11-15 | independent | Source ↗ |
| Muse Wonder | 68.1% | 2026-11-15 | independent | Source ↗ |
| Muse Spark Open | 49.7% | 2026-11-15 | independent | Source ↗ |
| Qwen3.8 27B | 48.2% | 2026-11-15 | independent | Source ↗ |
| K2 Horizon 3.7B | 14.7% | 2026-11-15 | independent | Source ↗ |
| Claude Opus 5.5 | 69.5% | 2026-11-15 | independent | Source ↗ |
| Command A+ | 1.8% | 2026-11-15 | independent | Source ↗ |
| Nemotron 3.5 Lightning | 0.8% | 2026-11-15 | independent | Source ↗ |
| Mistral Large 4 | 57.8% | 2026-11-15 | independent | Source ↗ |
Human Baseline & Difficulty Horizon
Calibrated human reference points, specialist benchmarks, and ceiling thresholds.No peer-reviewed human baseline has been published for this benchmark.
Metric & Scoring Methodology
Verification protocols, aggregation formulas, and specialized metric variants.task success (%)Dataset & Compute Cost
Evaluation volume, public availability, API pricing, and local hardware requirements.$5 – $20 USD for full benchmark evaluation run on frontier APIs.
1x NVIDIA RTX 4090 (24GB) or A100 (40GB/80GB) via vLLM / SGLang
How to Run & Reproduce
Standardized evaluation protocols, CLI commands, and reproducible runner templates.lm_eval --model hf --model_args pretrained=<model_path> --tasks automationbench-aa --batch_size autoopencompass --datasets automationbench-aa --models <model_config># Standard API Evaluation Loop
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="gpt-4o",
messages=[{"role": "user", "content": prompt}],
temperature=0.0,
)When publishing results for AutomationBench-AA, always report the exact prompt template, few-shot exemplar ordering, sampling temperature (temperature=0), maximum reasoning budget tokens, and the precise timestamped model snapshot ID.
Contamination & Memorization Analysis
Audit of pretraining exposure risks, memorization vectors, and refresh policies.Static fixed snapshot
Public on web / HuggingFace