Breaking

Gemini 3.8 Flash appears on Google Cloud quotas as Gemini app users report it in chats

September 2, 2026 at 09:56 EDT

Foundation Models
Software Dev & Coding
AI Agents

Google has not announced Gemini 3.8 Flash, but the model name has begun showing up on the Google Cloud Console quotas page, a step that has often come just before public availability, and some Gemini app conversations are already being attributed to the new Flash build. The listing is a pre-release signal rather than a documented product launch: official Google Cloud model pages still highlight Gemini 3.7 Flash, and no public model card, API ID, price sheet, or benchmark suite has been published for 3.8.

September 2, 2026 · Google

Gemini 3.8 Flash is in the console — not yet on the model card

The model name now appears on the Google Cloud quotas page, and some Gemini app chats already name it. Google has not published a 3.8 launch, price, or benchmark sheet.

Flash ship intervals (days)

~23

3.6 → 3.7
Jul 21 to Aug 13

~20

3.7 → 3.8 quotas
Aug 13 to Sep 2

Two Flash jumps in under a month — close to the monthly cadence Google has described.

Aug 13

Gemini 3.7 Flash public release

$0.75 / $3.75

3.7 intro in/out per 1M tokens — 3.8 rates unlisted

Preview

Jetski internal tests; some Gemini app chats

How the signal usually unfolds

1 Internal Jetski preview → 2 Cloud quotas listing → 3 Gemini app A/B chats → 4 Public model card still pending

What developers note

Quota listings have often preceded a Flash drop. Early testers called 3.8 noticeably better than 3.7 on Jetski, and some consumer chats already identify the new name.

What is still missing

No official 3.8 price, API ID, context limit, or benchmark. Cloud docs still feature 3.7 Flash. Hardcoded model IDs look fragile on a three-week cycle.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.