Gemini 3.5 Flash is the new default AI model from Google, designed to perform actions rather than merely provide answers.

      Unveiled at I/O 2026, the model is crafted to manage intricate, multi-step workflows, and is now operational across all of Google's consumer, developer, and enterprise platforms.

      Google has previously portrayed its Gemini Flash models as the quicker, more affordable alternative to its primary Pro tier. However, this narrative shifts with Gemini 3.5 Flash. Revealed during I/O 2026, the latest model surpasses Gemini 3.1 Pro in coding and agentic metrics, with Google asserting it achieves four times the speed of similar frontier models, “often at less than half the cost.”

      Designed for agents

      Gemini 3.5 Flash caters specifically to long-horizon agentic tasks, which involve workflows where AI must plan, construct, and iterate through various steps. Google claims it can perform tasks that previously took developers days or auditors weeks in a fraction of the usual time.

      Introducing Gemini 3.5 Flash — our most advanced agentic and coding model to date. It offers frontier-level performance at four times the speed of comparable models, often at less than half the cost. Available starting today. 🧵#GoogleIO pic.twitter.com/jLhqozutwG— Google (@Google) May 19, 2026

      Benchmark results include scores of 76.2% on Terminal-bench 2.1, an Elo rating of 1656 on GDPval-AA, and 83.6% on MCP Atlas. The model also achieved 84.2% on CharXiv Reasoning, which assesses multimodal understanding.

      The model utilizes Google’s Antigravity, an agent-centric development platform, to deploy numerous subagents simultaneously, allowing it to address more complex tasks. For consumers, 3.5 Flash now serves as the default model powering the Gemini app and AI Mode in Google Search.

      It also powers Gemini Spark, a newly announced personal AI agent by Google that operates continuously to perform actions on behalf of users. Currently, Google is rolling out Spark to select testers, with a wider beta expected for Google AI Ultra subscribers in the U.S. next week.

      Deployment and accessibility

      Gemini 3.5 Flash is currently available worldwide in the Gemini app and AI Mode in Search for consumers, as well as through Google AI Studio, the Gemini API, and Android Studio for developers. Enterprise users can access it through both the Gemini Enterprise Agent Platform and Gemini Enterprise. Google has also confirmed that Gemini 3.5 Pro is undergoing internal testing and is anticipated to launch next month.

      This release positions agentic capabilities at the forefront of Google’s AI strategy. Rather than solely competing on the accuracy of answers, Google is focusing on the next evolution of AI—developing models that can perform tasks like booking appointments, writing and executing code, managing workflows, and more, all with minimal user input. With Gemini 3.5 Pro set to launch next month, it will be intriguing to see how far Google intends to pursue this vision.

      Pranob is an experienced tech journalist with over eight years of expertise in consumer technology coverage. His work has been...

      Google Search is introducing AI agents that will keep an eye on the web for you.

      Set up an agent once, and Search will alert you when it discovers what you're seeking.

      During its I/O 2026 keynote, Google announced a significant revamp of Search, featuring AI agents, a redesigned search box, and agentic coding capabilities that can create custom apps and dashboards on the spot.

      A revamped search box

      Read more

      Google Pomelli can now develop your entire brand from the ground up.

      From zero to a complete brand kit and a functional website in just a few clicks.

      Google I/O 2026 heavily emphasizes empowering users to create their projects. In addition to the new Gemini 3.5 Flash model, a new Antigravity 2.0 IDE, and a new Google AI Studio, the company has also improved Google Pomelli, a tool designed to assist small businesses in establishing their brand identity.

      Initially launched last year through Google Labs, Google reports that Pomelli has already helped millions of businesses generate professional product images, social campaigns, and advertisements. Today, Google is introducing an array of new agentic features that expand the tool beyond just social media posts.

      Read more

      Google’s AI subscriptions are receiving a new $100 tier, alongside a price reduction and new features across all plans.

      Google announced a revamp of its AI subscriptions at I/O 2026, introducing a new AI Ultra tier, reducing the top plan's price from $250 to $200, and implementing new models and features across all subscription plans.

      New tiers and pricing

      Read more