What has changed with GPT-6 Astra? Basic specifications of the new model that supports 'thinking, researching, and creating' in ChatGPT
Unlike previous models, Astra organizes information, makes judgments, and chains processes like reading a document, researching gaps, and compiling a report into a single workflow. It includes a reasoning.effort setting (low to max) to adjust depth for tasks like comparing conditions or analyzing code, with higher effort increasing time and usage costs.
Key points
- GPT-6 Astra combines reasoning, browsing, coding, and document creation into multi-step workflows, not just text generation
- Reasoning depth is adjustable via **reasoning.effort** (low to max), affecting response time and usage costs for complex tasks
- Available only to **Pro, Business, Enterprise** (GPT-6 Pro) and **Plus** (via Work/Codex) users, with rollout dependent on location and app
Astra is not yet available to all ChatGPT users. As of September 2026, it powers GPT-6 Pro for Pro, Business, and Enterprise plans, while Plus users access it via ChatGPT Work and Codex. Availability depends on location, plan, and app environment. The model can browse the web, operate applications, and generate code—but users must verify outputs, permissions, and actions for accuracy, especially for critical tasks like deletions or purchases.
Model page: GPT-6 Astra →
The story so far
5 episodes →- What has changed with GPT-6 Astra? Basic specifications of the new model that supports 'thinking, researching, and creating' in ChatGPTthis story
What has changed with GPT-6 Astra? Basic specifications of the new model that supports 'thinking, researching, and creating' in ChatGPT
note.com · 21 September 2026
Loading the full article…
This text was published by note.com and written by 霧島さん. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
Coverage and discussion
2sourcesThe headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Generative AI & Models
All →- Anthropic releases Claude Opus 5.5; OpenAI halves prices for GPT-6 Sol and Luna · 14 src
- OpenAI releases ChatGPT Images 2.5 with sketch, point, and template tools · 4 src
- FreedomIntelligence releases HuatuoGPT-3 multimodal model on Hugging Face · 1 src
- PrismML launches 1-bit Bonsai LLM for Qualcomm smart glasses · 1 src
- BottleCap AI cuts Qwen3.8-27B’s reasoning tokens by 37.2% with ThinkingCap-Qwen3.8-27B · 1 src
Comments
via GitHub Discussions