Google launches Gemini 3.5 Flash, fastest agentic AI
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Google rolled out Gemini 3.5 Flash, described as its strongest agentic and coding model yet, for long‑horizon agentic tasks in the Gemini app and Search’s AI Mode.
- Gemini 3.5 Flash is billed as fast and capable of executing complex, agentic workflows, positioning it as the centerpiece of Google’s AI‑agent strategy announced at I/O 2026.
- The model costs roughly three times more per token than earlier Gemini versions and is token‑hungry, leading to run‑time expenses about twice those of Gemini 3.1 Pro on comparable tasks.
- Google claims the new model can slash enterprise AI spending by over $1 billion annually, highlighting its cost‑saving potential despite higher per‑token pricing.
- Gemini 3.5 Flash is paired with new AI agents such as Gemini Spark and Gemini Omni, expanding Google’s focus from chatbots to agentic applications across search, coding, and scientific tools.
Why it matters: Enterprises stand to save over $1 billion annually as Google touts Gemini 3.5 Flash’s efficiency, while developers and businesses must absorb its three‑fold per‑token price, making high‑volume usage costlier than previous Gemini models and shifting the cost‑benefit balance toward agentic applications, potentially reshaping AI budgeting priorities.
