Gemini 3.7 Flash is out. We already build with Gemini.

Google released Gemini 3.7 Flash on 13 August 2026. They call it their most intelligent workhorse model for coding and agents. Introductory API price: $0.75 per million input tokens and $3.75 per million output tokens through 31 December 2026. (Google, Lovable changelog)

Lovable moved it in the same window. New AI features in a Lovable app now default to Gemini 3.7 Flash unless you named another model.

Why it is on our journal

We use Claude and Gemini to think, Lovable plus Supabase to ship. That is the split in how we build.

A faster cheap Gemini model means two practical things. Quicker studio drafts. And a client app with its own AI feature can use it without a separate Google project.

We still use Claude for a longer careful pass. We do not swap the whole studio every launch.

What we will not claim

We will not claim 3.7 Flash replaces developers. Speed and price are the story. The last 30% is still auth, payments, domain, and email.

Tell us the problem.