Gemini 3.7 Flash Launches at Half the Price of 3.6 Flash — But Only Until December 31

A pricing figure for a model called Gemini 3.7 Flash started circulating on X on August 13, claiming a launch rate of $0.75 per million input tokens and $3.75 per million output tokens, alongside claims that it would ship "in a few hours." Hours later, it did — and the leaked number was exactly right.

Google officially launched Gemini 3.7 Flash on August 13, confirming the $0.75/$3.75 rate directly in its announcement. The catch that didn't make it into most of the leak chatter: this is introductory pricing with a defined expiration date, not a permanent price cut. It runs through December 31, 2026, after which the rate doubles to $1.50/$7.50 per million tokens — the same price Gemini 3.6 Flash launched at.

01What Actually Shipped

Gemini 3.7 Flash arrived just three weeks after Gemini 3.6 Flash, continuing Google's rapid Flash-tier release cadence. It's positioned for coding and autonomous agent work, with Google citing benchmark gains over 3.6 Flash: the DeepSWE v1.1 debugging benchmark moved from 49.0% to 65.3%, and FrontierCode 1.1 Main improved from 34.4% to 43.6%. The model accepts text, images, audio, and video across a 1-million-token context window, and now powers Gemini Spark, Google's persistent AI agent available to AI Pro and Ultra subscribers.

02The Part Worth Budgeting Around

A halved price against 3.6 Flash's real launch rate is a meaningfully steeper cut than Google's last two Flash releases delivered — 3.6 Flash itself only moved output pricing down 17% against the 3.5 Flash it replaced. That's a fair amount of margin for Google to be temporarily giving up, and the December 31 expiration date is the clue as to why: this reads as an adoption push, not a permanent repricing.

Key takeaway

If you build a production workload on Gemini 3.7 Flash today, the unit economics you're seeing right now aren't the unit economics you'll have on January 1. Budget for the $1.50/$7.50 rate from the start if you're planning anything beyond a short evaluation window, so the transition isn't a surprise.

03What Else Is Going On

Gemini 3.5 Pro, Google's flagship-tier model, remains unshipped as of this launch — extending its delay into a fourth consecutive month with no timeline given. Three Flash-tier releases in six weeks against a stalled Pro tier suggests Google is currently defending the high-volume, cost-sensitive segment of the market rather than chasing frontier reasoning parity right now. Worth knowing if you're waiting on Pro specifically rather than evaluating Flash.

04What to Do Right Now

Run your real token volumes through the calculator at both the introductory and post-expiration rates, so you have both numbers before you commit to anything long-running. If your workload is a short-term evaluation or a project that wraps before year-end, the $0.75/$3.75 rate is straightforwardly good. If it's an ongoing production system, plan around $1.50/$7.50 as the real number and treat anything below that as a temporary bonus rather than your baseline.