Skip to content
Google· Launch

Early Gemini 4 Pro tests leaked – model yet to be released

Internal checkpoints for Google DeepMind's upcoming AI model, Gemini 4 Pro, have been discovered by developers. Unofficial tests indicate strong performance in physics simulation and coding, though the model has not yet been officially launched.

By the Aheadline editorial team·29 sep. 2026·2 min read·Source: Entity-watch: Google DeepMindVerifierad signalAI-generated
Early Gemini 4 Pro tests leaked – model yet to be released
Early Gemini 4 Pro tests leaked – model yet to be released
Early Gemini 4 Pro tests leaked – model yet to be released
By · Verktygs- & infrastrukturreporter
Last updated
Vad betyder det för mig?

What happened?

Google DeepMind is preparing its next general AI model, Gemini 4 Pro. Developers have identified new internal checkpoints under the codename "barium-b" (Checkpoint 2) and conducted unofficial tests. The model reportedly demonstrates strong results in 3D physics simulation, code generation, and long-context reasoning.

Key facts

Kodnamnbarium-b (Checkpoint 2)
UtvecklareGoogle DeepMind
StatusEj officiellt lanserad

Why it matters

The leak provides an indication of what the next generation of Google's large-scale AI models may offer. As the models remain in development, the tests primarily highlight how manufacturers evaluate physics comprehension and coding ability within internal test environments.

Who is affected?

This information is primarily of interest to AI developers, software researchers, and technology companies monitoring the development of large language models. As the model has not been officially launched, typical end-users remain unaffected.

Impact on the EU

It is not yet clear how Gemini 4 Pro will be deployed in the EU or how it complies with the requirements of the EU AI Act regarding transparency and risk assessment for advanced AI models.

What else you should know

These reports are based on tests of leaked internal names and checkpoints. There are currently no official benchmarks from independent parties or confirmation from Google regarding its final performance relative to competing models.

Frequently asked questions

Quick answers about this story

Vad har hänt?
Utvecklare har upptäckt interna kontrollpunkter ("barium-b") för Googles kommande AI-modell Gemini 4 Pro och utfört tidiga inofficiella tester.
När hände det?
Informationen och de inofficiella testresultaten publicerades den 27 september 2026.
Varför spelar det roll?
Det ger en första indikation på kapaciteten hos Googles nästa modellgeneration inom fysiksimulering och kodning, även om modellen ännu inte är officiellt lanserad.
Är Gemini 4 Pro tillgänglig för allmänheten?
Nej, modellen är under utveckling och har inte lanserats officiellt av Google DeepMind.
Original source
Entity-watch: Google DeepMind·eu.36kr.com

The link opens in a new window and leads to the publisher's own site.

Verifierad signal

Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.

AI-verktyg i artikeln

Topics

#Google DeepMind#Gemini
[ STAY UP TO DATE ]

Get similar news straight to your inbox

No affiliate linksCancel anytimeGDPR-friendly
[ Frequency ]
[ What do you want to read about? ]

You'll receive updates on 2 topics.

The reader's room

Send in a question or an addition. The newsroom reads everything before it's published and replies when relevant. No AI-generated text – just people.

Sign in to submit a comment or question.

Loading comments…
How this affects you

Read the article through your role

  • Decide whether this affects strategy over 6–12 months or is just noise.
  • Discuss with leadership: do we own the right question or does ownership need to move?
  • Ask: what risk are we taking by NOT acting on this this quarter?

Generated angle — not editorial analysis of "Early Gemini 4 Pro tests leaked – model yet to be released"