Agent SaaS
Google's Fast AI Model Completes Nearly Two-Thirds of Business Tasks
Agentic SaaS workflows, compared with Kimi K3 and GPT-5.6 Sol
Google's Fast AI Model Completes Nearly Two-Thirds of Business Tasks
Why it matters: Artificial Analysis scored Gemini 3.7 Flash (high) at 62.7% on AutomationBench-AA. That is 9.7 points above Kimi K3 (max, 53%) and 11.5 points above GPT-5.6 Sol (max, 51.2%) for Google's third Flash release in three months.
View source →- Source
- Artificial Analysis
- Unit
- AutomationBench-AA (%)
- Data as of
- August 13, 2026
- Scope
- Google Gemini 3.7 Flash versus Kimi K3 and OpenAI GPT-5.6 Sol