AutomationBench-AA - Finance - Raw Objective Completion: leaderboard
Artificial Analysis's implementation of Zapier's AutomationBench restricted to its Finance tasks: the share of task objectives an agent completes in simulated SaaS applications, counted before the guardrail-violation adjustment used by the AutomationBench-AA headline. Reported 0-100 (%); higher is better.
Metric: Raw Objective Completion (%). Source: artificialanalysis.ai. Saturation forecast: Around February 2027. 243 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 4 Carbon (High) | 97.34 | #1 (Gemini 4 Carbon) |
| 2 | Gemini 4 Argon (High) | 94.62 | #9 (Gemini 4 Argon) |
| 3 | MiniMax-M3.1-Flash | 93.72 | #52 |
| 4 | Grok 4.7 (xHigh) | 93.49 | #26 (Grok 4.7) |
| 5 | Kimi K3.1 Pro | 93.32 | #17 |
| 6 | Grok 4.7 (High) | 92.19 | #26 (Grok 4.7) |
| 7 | Grok 4.6 (xHigh) | 92.16 | #29 (Grok 4.6) |
| 8 | DeepSeek V4.1 Flash (1017) (High) | 91.96 | #27 (DeepSeek V4.1 Flash (1017)) |
| 9 | Grok 4.6 (High) | 91.74 | #29 (Grok 4.6) |
| 10 | GLM-5.3 (Max) | 91.17 | #44 (GLM-5.3) |
| 11 | Qwen3.8 2.4T A95B | 90.52 | #59 |
| 12 | Ling 3.1 Flash | 90.3 | #75 |
| 13 | DeepSeek V4.1 Flash (Max) | 89.84 | #37 (DeepSeek V4.1 Flash) |
| 14 | GPT-6.1 Sol (xHigh) | 89.3 | #7 (GPT-6.1 Sol) |
| 15 | Grok 4.6 (Medium) | 89.27 | #29 (Grok 4.6) |
No result here: #3 Claude Opus 5.5, #10 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=automationbench-aa-finance-raw-objective-completion · How It Works · Data refreshed daily, snapshot 2026-11-05.