Photo: Merlin Lightpainting / Pexels
Just three weeks after launching Gemini 3.7 Flash, Google is already rolling out its successor in an aggressive push to dominate the developer AI market
Google DeepMind is set to publicly reveal Gemini 3.8 Flash on Wednesday, September 2, continuing what has become an almost absurdly fast cadence of AI model releases from the search giant. The model, internally codenamed “skimaki,” has already completed production deployment after testing on Google’s Jetski coding platform throughout August.
To put the pace in perspective: Gemini 3.7 Flash launched on August 13, 2026. That was itself just three weeks after version 3.6 Flash.
What’s actually changing #
Early tester feedback points to Gemini 3.8 Flash as more of a refinement than a revolution. The primary improvements center on reducing verbose outputs, a persistent complaint with earlier Flash models. The model also reportedly addresses specific issues identified in Gemini 3.7 Flash.
CEO Sundar Pichai has indicated Google is targeting a release cadence of roughly one new model per month.
The pricing play #
Gemini 3.7 Flash debuted with introductory pricing of $0.75 per million input tokens and $3.75 per million output tokens, rates that hold through year-end 2026.
Why the speed matters #
Google’s Jetski coding platform, where 3.8 Flash Preview testing took place, appears to serve as the proving ground for these rapid iterations. By dogfooding new models internally on coding tasks before public release, Google gets real-world performance data while simultaneously improving its own developer tools.
The leaker community, which flagged the “skimaki” codename and Wednesday reveal date, has become an increasingly reliable early-warning system for Google’s AI releases.
Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our