Particle.news
Download on the App Store

Google Releases Gemini 3.7 Flash for Coding and Agent Workflows

The model is meant to capture developer and agent workloads with high context and low cost under new EU transparency rules.

Overview

  • Gemini 3.7 Flash, which Google released on August 13, is the latest low‑latency 'Flash' model targeted at software engineering and automated business tasks.
  • The model accepts multimodal inputs and supports a context window exceeding one million tokens so it can process long codebases, documents, or mixed media in a single session.
  • Google says 3.7 Flash improves coding efficiency versus the 3.6 Flash release and can drive agent pipelines that combine Antigravity with Nano Banana to produce playable game assets from a text prompt.
  • Google is offering an introductory price of $0.75 per million input tokens and $3.75 per million output tokens through December 31, 2026, and is making the model available via the Gemini API, Google AI Studio, Antigravity, and Gemini Spark.
  • The rollout aligns with the EU AI Act transparency rules and Google’s SynthID watermarking commitments, while a higher‑end Pro flagship remains in testing as Alphabet balances heavy AI infrastructure spending.