Google announced Gemini 3.5 Flash at its I/O 2026 developer conference, positioning it as the first model in a new Gemini 3.5 series that the company says combines frontier intelligence with agentic action. <cite index="3-7,3-8">Google launched Gemini 3.5 Flash as the first in its latest series of models combining frontier intelligence with action.</cite> <cite index="3-9,3-10">Gemini 3.5 Flash is generally available via Google's agent-first development platform Google Antigravity, the Gemini API in Google AI Studio, and Android Studio.</cite>
Benchmarks and positioning
The headline claim inverts the company's usual model hierarchy: a Flash-tier (speed-optimized) model outperforming the previous Pro-tier (capability-optimized) release. <cite index="3-12,3-13">Google says Gemini 3.5 Flash delivers intelligence that rivals large flagship models at speeds typical of the Flash series, and outperforms Gemini 3.1 Pro on coding and agentic benchmarks including Terminal-Bench 2.1 (76.2%), GDPval-AA (1656 Elo) and MCP Atlas (83.6%).</cite> <cite index="1-15,1-16">It scores 84.2% on CharXiv Reasoning, which tests multimodal understanding.</cite>
<cite index="3-26">Google places the model in the top-right quadrant of the Artificial Analysis index, arguing it delivers frontier-level intelligence at exceptional speed.</cite> <cite index="5-9">The company says 3.5 Flash delivers four times the output token generation speed of competing frontier models.</cite> According to Google's published comparisons, the model trails its predecessor Gemini 3.1 Pro on long-context retrieval (MRCR v2 at 128k tokens) and Humanity's Last Exam, areas where raw knowledge depth weighs more heavily than agentic tool use.
Default in AI Mode Search
Google is rolling the model into its consumer surfaces immediately. <cite index="3-19,3-20,3-21">AI Mode, Google's most powerful AI Search experience, has surpassed one billion monthly users, and is being upgraded with Gemini 3.5 Flash as the new default model globally; Google says AI Mode queries have more than doubled every quarter since launch.</cite> The company also introduced a redesigned Search box and persistent "information agents" that run continuously to monitor topics for users.
Developer platform and Antigravity 2.0
Alongside the model, Google shipped an updated agentic IDE. <cite index="5-11,5-12">Google updated Antigravity, its agentic development platform, to version 2.0, with an Editor view that functions like a familiar IDE interface with an agent sidebar, and a Manager view that serves as a control center for orchestrating multiple agents working in parallel across workspaces.</cite> <cite index="5-13">Google claims a hyper-optimized version of Gemini 3.5 Flash inside Antigravity reaches significantly faster generation speeds than standard frontier model deployments.</cite>
Enterprise adoption was highlighted as part of the launch. <cite index="1-26,1-27,1-28,1-29,1-30,1-31,1-32,1-33,1-34">Shopify is running subagents in parallel for data analysis to power merchant growth forecasts; Macquarie Bank is piloting the model for customer onboarding, reasoning over complex 100+ page documents to retrieve information and make recommendations; Salesforce is integrating 3.5 Flash into Agentforce to automate enterprise tasks using multiple subagents that retain context across multi-turn tool calling; and Ramp is using it for invoice OCR.</cite>
Scale and what's next
Google framed the release against expanding infrastructure usage. <cite index="5-10">CEO Sundar Pichai said Google is now processing more than 3.2 quadrillion tokens per month, up from 480 trillion tokens per month at I/O 2025.</cite> <cite index="3-33,3-34">Google said it is also working on Gemini 3.5 Pro, which is being used internally and is expected to roll out next month.</cite> Until then, Gemini 3.1 Pro remains the recommended option for the deepest pure-reasoning tasks, while 3.5 Flash assumes default-model duty across the Gemini app, AI Mode, and Google's developer surfaces.