Gemini 3.5 Flash
Google · released 20 May 2026 · proprietary API
Flagship-class fast model; 280+ tok/s, default in Gemini app + Search AI Mode.
Context window
1M
1,000,000 tokens
Input price
$1.50
per 1M tokens
Output price
$9
per 1M tokens
Blended
$3.38
3:1 in:out mix
Weights
Proprietary
API only
Released
20 May 2026
Where it ranks
Of the 16 models tracked, Gemini 3.5 Flash is #10 cheapest to run (blended $/1M at a 3:1 input:output mix), #8 by input price, and #2 (tied) by context window.
See the full Frontier Index →Compare with
The nearest models by blended cost to run.
Put it to work
Specs from BitByteCore's hand-verified Frontier Ledger (v2026-07-04, verified 21 Jun 2026). Model pricing moves fast — confirm with Google before you rely on a figure.