Google moved Gemini 3.5 Flash to general availability, opening the new Flash-class model to developers through the Gemini API in Google AI Studio, Vertex AI, Android Studio, and the company's Antigravity agent platform. The release lands the same week as Microsoft Build 2026, intensifying the head-to-head between the two companies' AI stacks.
Gemini 3.5 Flash is the first in a new Flash series Google describes as combining "frontier intelligence with action." In Google's published benchmarks, the model outperforms Gemini 3.1 Pro on a set of agentic and coding tasks while delivering Flash-class latency and pricing. Google's positioning is unsubtle: developers who were defaulting to a Pro-class model for quality should now reach for Flash for most agent and tool-use workloads.
The launch is tightly bundled with Antigravity, Google's agent-building platform, where 3.5 Flash becomes the default reasoning model for new agents. The combination of low latency, native tool use, and direct integration with Google Workspace, Search, and Maps APIs is aimed squarely at the same enterprise agent buyer Microsoft is targeting with Azure Agent Mesh.
Google also previewed Gemini 3.1 Ultra alongside the GA launch, with a 2-million token context window that handles text, image, audio, and video natively without intermediate transcription. Ultra remains in limited preview while Flash carries the day-one developer story.
Pricing was not the announcement's headline, but Google did confirm Flash-class economics: input and output tokens are priced well below Pro-class peers, and Antigravity-hosted agents get a free monthly allowance. For developers, the practical takeaway is that the cost-quality frontier moved again, and a meaningful share of current Pro-tier workloads are now better served on 3.5 Flash.
Source: [BuildFastWithAI](https://www.buildfastwithai.com/blogs/ai-news-today-june-2-2026)