GEMINI LABJP
FLASH36 — Gemini 3.6 Flash arrived on July 21, consuming 17% fewer output tokens than 3.5 Flash and priced at $1.50 input and $7.50 output per 1M tokensSTEPS — 3.6 Flash takes fewer reasoning steps and tool calls to finish multi-step workflows, which shows up as lower spend on agentic runsLITE — Gemini 3.5 Flash-Lite targets high throughput and low latency at $0.30 input and $2.50 output per 1M tokens, aimed at agentic search and document processing at volumeCYBER — Google announced 3.5 Flash Cyber alongside 3.6 Flash and 3.5 Flash-Lite, and teased Gemini 4WHERE — Both 3.6 Flash and 3.5 Flash-Lite are available through Google Antigravity, AI Studio, and Android StudioHOME — Gemini for Home now holds conversational context for 15 minutes, and Gemini Live reached the first-generation Google Home Mini and Nest HubFLASH36 — Gemini 3.6 Flash arrived on July 21, consuming 17% fewer output tokens than 3.5 Flash and priced at $1.50 input and $7.50 output per 1M tokensSTEPS — 3.6 Flash takes fewer reasoning steps and tool calls to finish multi-step workflows, which shows up as lower spend on agentic runsLITE — Gemini 3.5 Flash-Lite targets high throughput and low latency at $0.30 input and $2.50 output per 1M tokens, aimed at agentic search and document processing at volumeCYBER — Google announced 3.5 Flash Cyber alongside 3.6 Flash and 3.5 Flash-Lite, and teased Gemini 4WHERE — Both 3.6 Flash and 3.5 Flash-Lite are available through Google Antigravity, AI Studio, and Android StudioHOME — Gemini for Home now holds conversational context for 15 minutes, and Gemini Live reached the first-generation Google Home Mini and Nest Hub
TAG

Cost Modeling

1 articles
Back to all tags
Related:
Gemini API1Flash-Lite1Model Routing1Indie Development1
Gemini API/2026-07-30Advanced

What Decided Our Cascade's Economics Wasn't the Escalation Rate

Putting Flash-Lite in front of 3.6 Flash and escalating only the hard items looks like an easy win. Estimating it from the escalation rate alone will mislead you. Here is the break-even solved with the escalated subset's output length in the equation.