Google Drops Gemini Flash Output Cost

What changed

On June 10, 2026, Google reduced Gemini 3.5 Flash output pricing from $0.75 to $0.60 per million tokens. That is a 20% cut on output, while input pricing remains unchanged at $1.50/MTok.

The current registry shows Gemini 3.5 Flash output at $9.00/MTok after a subsequent pricing restructure. This article covers the June 2026 change as recorded in the pricing changelog at that time.

Why it matters

Gemini 3.5 Flash ranks first for classification at 92/100 confidence and extraction at 90/100 confidence. It handles high-volume structured tasks at a fraction of what flagship models charge per call.

For a classification pipeline running 100,000 calls/day at 500 input and 200 output tokens:

• -Input cost: 100,000 x 500 / 1,000,000 x $1.50 = $75/day
• -Output cost (old): 100,000 x 200 / 1,000,000 x $0.75 = $15/day
• -Output cost (new): 100,000 x 200 / 1,000,000 x $0.60 = $12/day

Per-call savings look small, but at 100,000 calls per day the output cut saves ~$90/month.

Google's cost advantage in the fast tier

Google now offers four models priced under $2/MTok input across the fast tier:

• -Gemini 3.5 Flash-Lite: $0.25/$1.50
• -Gemini 3 Flash: $0.50/$3.00
• -Claude Haiku 4.5: $1.00/$5.00 (Anthropic)
• -Gemini 3.5 Flash: $1.50/$9.00

For workloads where Flash-tier confidence scores meet the bar, Google models are consistently cheapest.

What to do

If you run classification or extraction workloads on a flagship model, test Gemini 3.5 Flash instead. At 92/100 confidence for classification, Flash competes with models that cost 20x more per call.

Run a free cost comparison at /teardown, or start tracking your real spend with a free dashboard.