Google Introduces Gemini 3.5 Flash: Focus on Agentic Workflows and Coding

ArticleKoray Kavukcuoglu, Jeff Dean, Oriol Vinyals, Noam ShazeerAnnouncement

Google has introduced the Gemini 3.5 model family, engineered to execute complex, agentic workflows and software development tasks. The series debuts with Gemini 3.5 Flash, offering frontier-level performance at high inference speeds, generating output tokens four times faster than competing frontier models. Alongside the release, Google detailed enterprise adoption, the upcoming Gemini 3.5 Pro, and a 24/7 personal assistant called Gemini Spark.

Key points

  1. Gemini 3.5 Flash is available globally across the Gemini app, Search AI Mode, Google Antigravity, Gemini API (AI Studio and Android Studio), and Gemini Enterprise.
  2. The model scores 76.2% on Terminal-Bench 2.1, 1656 Elo on GDPval-AA, 83.6% on MCP Atlas, and 84.2% on CharXiv Reasoning, with 4x faster token output than other frontier models.
  3. Gemini 3.5 Pro is currently in internal testing and scheduled for rollout next month.
  4. Coupled with the Google Antigravity harness, the model orchestrates collaborative subagents for multi-step tasks such as codebase migration to Next.js and automated UI generation.
  5. Enterprise partners including Shopify, Salesforce (Agentforce), Databricks, Macquarie Bank, Ramp, and Xero are deploying 3.5 Flash for data analysis, customer onboarding, and accounting workflows.
  6. Gemini Spark, an always-on personal agent powered by 3.5 Flash, is launching to trusted testers and US Google AI Ultra subscribers.

Why it matters

Gemini 3.5 Flash signifies a major transition towards action-oriented, autonomous agents capable of handling multi-day tasks in minutes. By eliminating the trade-off between latency, cost, and high-reasoning capabilities, it accelerates practical agentic deployment across software engineering and enterprise operations.

Analysis

Google is positioning its Flash tier as a first-class agentic runtime rather than merely a lightweight fallback. Combining subagent orchestration via Antigravity with high token throughput addresses the primary bottleneck of autonomous workflows: compounding latency and cost. The early integrations with Salesforce Agentforce and Databricks demonstrate an immediate enterprise push into autonomous multi-turn tool calling.

Found in

Summary generated by AI (Gemini 3.8 Flash, 27 September 2026). It may contain errors – the original source is authoritative.

Related entries

Similar in content, found by the AI search.

  • Link:Google DeepMind

    Google DeepMind Announces Gemini 3.8 Flash for Agentic Workflows

    Google DeepMind has introduced Gemini 3.8 Flash, an agentic workhorse model designed for complex multi-step reasoning, coding, and knowledge tasks at low latency. The multimodal model supports a 1M token input context alongside extensive tool integration.

    AI & LLMs· Announcement

  • Article:Tulsee Doshi

    Google introduces Gemini 3.7 Flash

    Google has announced Gemini 3.7 Flash, an updated workhorse model focused on software engineering and autonomous agents. Arriving three weeks after version 3.6 Flash, the model delivers measurable benchmark gains alongside a halved introductory token price.

    AI & LLMs· Announcement

  • Article:Tulsee Doshi

    Google Announces Gemini 3 Flash

    Google has launched Gemini 3 Flash, a cost-effective and high-speed model designed to bring Gemini 3 reasoning capabilities to latency-sensitive and high-frequency workflows.

    AI & LLMs· Announcement

  • Article:Tulsee Doshi, Raluca Ada Popa

    Google Introduces Gemini 3.8 Flash and 3.8 Flash Cyber

    Google has announced Gemini 3.8 Flash and Gemini 3.8 Flash Cyber, targeting long-horizon coding, agentic workflows, and specialized defensive cybersecurity.

    AI & LLMs· Announcement

Back to the knowledge base

Let's talk about your project

Get in touch

Locations

  • Mattersburg
    Johann Nepomuk Bergerstraße 7/2/14
    7210 Mattersburg, Austria
  • Vienna
    Ungargasse 64-66/3/404
    1030 Wien, Austria

Parts of this content were created with the assistance of AI.