Google releases Gemini 3.7 Flash with big coding gains and a half-price launch offer

Google has released Gemini 3.7 Flash, a new version of its workhorse AI model built for coding and agent tasks, just three weeks after Gemini 3.6 Flash. It is available now, and Google is charging half the usual rate for it through the end of 2026.

This is not a new model built from scratch. Google describes 3.7 Flash as a refinement of 3.6 Flash, with algorithmic improvements to its reasoning rather than a fresh training run. The short gap between releases is the point: Google says developer feedback and those quicker improvements are driving a faster release rhythm for the Flash line.

Where the model actually improved

Google’s own benchmark numbers concentrate the gains in three areas: software engineering, web development, and document-heavy work.

Benchmark	Gemini 3.6 Flash	Gemini 3.7 Flash
FrontierCode 1.1 Main (production code)	34.4%	43.6%
DeepSWE v1.1 (long software tasks)	49.0%	65.3%
WebDev Arena (Elo)	1538	1588
GDP.pdf (document comprehension)	22.0%	34.0%
AutomationBench (workflow automation)	17.0%	30.4%

The rest of the spec sheet carries over what Flash users expect: a 1M-token context window, up to 64K output tokens, and support for text, image, audio, and video input. Developers can also tune how much the model thinks before answering, trading quality against cost and speed.

Your Complete iOS 26 Playbook:

Get our exclusive Ultimate iOS 26 Guide 📚 — absolutely FREE when you sign up for our newsletter below.

Pricing starts at half rate, but not forever

Through December 31, 2026, Gemini 3.7 Flash costs $0.75 per million input tokens and $3.75 per million output tokens via the API. That is half of what 3.6 Flash launched at. From January 1, 2027, the price doubles to $1.50 and $7.50.

If you are building on the API, that deadline matters more than the launch itself. Anything you ship on the introductory rate will cost twice as much to run in the new year.

Where you can use it

For regular Gemini users, the model shows up first in Gemini Spark, the background agent inside the Gemini app, for AI Pro and Ultra subscribers in over 160 countries. Google has not said whether 3.7 Flash becomes the default model for normal Gemini app chats.

Developers get it today in Google AI Studio, Android Studio, and Google Antigravity, and businesses through the Gemini Enterprise Agent Platform and the Gemini Enterprise app.

The flagship is still missing

The Flash line is moving fast, but Google’s top model is not. Bloomberg reports that Gemini 3.5 Pro remains delayed, with no release date from Google. Back at I/O 2026, Google said 3.5 Pro would arrive within a month of the event. Nearly two months later, the smaller Flash models keep shipping while the flagship stays internal.