Skip to content
OpenAI· Launch

OpenAI unveils GPT-6 Astra with new reasoning architecture

OpenAI has announced GPT-6 Astra, a model featuring latent recurrent reasoning and a context window of over one million tokens.

By the Aheadline editorial team·27 sep. 2026·2 min read·Source: Entity-watch: OpenAIVerifierad signalAI-generated
OpenAI unveils GPT-6 Astra with new reasoning architecture
OpenAI unveils GPT-6 Astra with new reasoning architecture
OpenAI unveils GPT-6 Astra with new reasoning architecture
By · Verktygs- & infrastrukturreporter
Vad betyder det för mig?

What happened?

OpenAI has announced GPT-6 Astra, a new AI model focused on advanced reasoning, software engineering, and data security. The model is built on a "neuralese recurrence" architecture, meaning deeper reasoning is performed directly within latent vector spaces. GPT-6 Astra has a context window of 1,050,000 tokens and supports a maximum output size of 128,000 tokens per request.

Key facts

Kontextfönster1 050 000 tokens
Maximal utdata128 000 tokens
Publiceringsdatum26 september 2026

Why it matters

By using recurrent depth, or a "looped transformer" architecture, the model can increase its computational capacity and perform more complex reasoning without the number of parameters scaling linearly. This aims to streamline processing for advanced problem-solving and code analysis.

Who is affected?

The model is primarily aimed at software developers, cybersecurity experts, and companies building complex AI applications via the OpenAI API. Developers are provided with controllable reasoning levels that can be adjusted as required.

Impact on the EU

Launch status for the EU has not yet been detailed separately by OpenAI. As previous model releases have been affected by the EU AI Act and GDPR, it remains unclear whether all features will be available to EU users simultaneously.

What else you should know

According to the source TheNextGenTechInsider, the model introduces a "looped transformer" architecture that enables increased computational depth interaction without a linear increase in the number of parameters. Developers are also given the ability to adjust the degree of reasoning in five distinct levels, from low to max.

Frequently asked questions

Quick answers about this story

Vad har hänt?
OpenAI tillkännagav sin nya modell GPT-6 Astra, inriktad på avancerade resonemang och datasäkerhet.
När hände det?
Modellen rapporterades och offentliggjordes den 26 september 2026.
Varför spelar det roll?
Modellen introducerar ny arkitektur med återkommande djup vilket tillåter mer komplexa resonemang i latenta vektorrum utan linjär ökning i parametrar.
Vilka berörs av nyheten?
Den riktar sig främst till mjukvaruutvecklare och cybersäkerhetsexperter som använder OpenAI:s API.
Original source
Entity-watch: OpenAI·thenextgentechinsider.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

#Safety#AI Safety#OpenAI#Models
[ 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 "OpenAI unveils GPT-6 Astra with new reasoning architecture"