产品更新

Google Already Testing Gemini 3.8 Flash Preview Just Weeks After 3.7 Release

Brave/androidheadlines.com · 2026年8月29日 02:15 · 阅读约 5 分钟
Google Already Testing Gemini 3.8 Flash Preview Just Weeks After 3.7 Release

Google Already Testing Gemini 3.8 Flash Preview Just Weeks After 3.7 Release

摘要: Trusted By Millions

facebook

+1.7m

X_logo_2023_original

+446.8K

+92.5k

instagram

linkedin2

Independ

正文

Trusted By Millions

facebook

+1.7m

X_logo_2023_original

+446.8K

+92.5k

instagram

linkedin2

Independent, expert Android News you can trust, since 2010. Publishing more Android than anyone else.

This Month

All Time

Newsletter

Stay Up to date with the latest Android News

Sign up to receive the latest

Android News

every weekday:

Only send updates once a week

Sign up!

Leave this field empty if you're human:

Android Headlines

/

Tech News

/

Artificial Intelligence News

/

Google Already Testing Gemini 3.8 Flash Preview Just Weeks After 3.7 Release

Google Already Testing Gemini 3.8 Flash Preview Just Weeks After 3.7 Release

Jean Leon

|

Aug 28, 2026

Add Android Headlines as a preferred source on Google

Google is reportedly testing Gemini 3.8 Flash internally on its Jetski engineering platform, just two weeks after launching Gemini 3.7 Flash on August 13, 2026. Reports from Nokia Power User, Business Insider, and AI leakers indicate the update targets improved agentic workflows, better code synthesis, and reduced hallucinations.

is moving at breakneck speed with its AI models right now. Barely two weeks after launching

Gemini 3.7 Flash

on August 13, 2026, reports reveal that Google testing

3.8 Flash is already happening behind closed doors.

First reported by

Business Insider

, Google staff are actively dogfooding an early 3.8 Flash preview on Jetski, the company’s internal coding platform. Engineers are putting the unannounced build through its paces against live coding tasks before opening up external APIs.

Smoother tool execution and paper trail clues

While Google hasn’t released official benchmarks or model cards yet, early feedback hints at noticeable performance jumps over 3.7. The internal push centers on fixing common headaches in lightweight models. The team is specifically sharpening multi-step agentic workflows, speeding up code


来源: Brave/androidheadlines.com

AI人工智能Google
阅读原文 · BRAVE/ANDROIDHEADLINES.COM →