GEMINI LABJP
CHAT — Ask Gemini in Chat goes live today, August 26, turning Google Chat into a single command line for search, drafting, catching up, and task managementLIMITS — Ask Gemini in Chat comes with promotional higher limits through October 1, after which standard usage limits applySHEETS — Sheets canvas turns a spreadsheet into an interactive, read-write application from a plain-language promptMEET — You can now start a Gemini note-taking session straight from the Google Meet home screen, including for in-person meetings, with the summary, action items, and full transcript saved to a Google DocMODELS — Gemini 3.7 Flash reached general availability on August 13, with introductory pricing running through December 31, 2026DEPRECATION — gemini-robotics-er-1.6-preview shuts down on August 31, five days from now. The ER 2 endpoints have been in public preview since July 30CHAT — Ask Gemini in Chat goes live today, August 26, turning Google Chat into a single command line for search, drafting, catching up, and task managementLIMITS — Ask Gemini in Chat comes with promotional higher limits through October 1, after which standard usage limits applySHEETS — Sheets canvas turns a spreadsheet into an interactive, read-write application from a plain-language promptMEET — You can now start a Gemini note-taking session straight from the Google Meet home screen, including for in-person meetings, with the summary, action items, and full transcript saved to a Google DocMODELS — Gemini 3.7 Flash reached general availability on August 13, with introductory pricing running through December 31, 2026DEPRECATION — gemini-robotics-er-1.6-preview shuts down on August 31, five days from now. The ER 2 endpoints have been in public preview since July 30
TAG

thinking-level

1 articles
Back to all tags
Related:
gemini1latency1cost1mobile1
Gemini Advanced/2026-07-12Advanced

Spend Deep Reasoning Only Where It's Needed: Per-Request thinking_level Routing in Gemini

Running every request at high thinking_level bloats latency and cost; forcing low drops accuracy on hard questions. This walks through a router that picks Gemini 3.x thinking_level per request from an inexpensive difficulty estimate, keeping p95 latency inside a mobile budget while reserving deep reasoning for the questions that need it — with measured numbers and working code.