GPT-6 Astra

6 Astra · Family: GPT

OpenAI's flagship GPT-family model, successor to GPT-5.6 Sol, focused on computer use, coding and reasoning. Announced September 4, 2026.

✓ Active⏳ Limited accessLLMMultimodalReasoning modelTool-using model📁 GPT

Release date

4 September 2026

🏢OpenAIProducer

Access:APIHostedDeployment:☁ Cloud

Overview

GPT-6 Astra is OpenAI's flagship AI model from the GPT family, announced on September 4, 2026 as the successor to GPT-5.6 Sol. OpenAI describes it as a model focused on computer use, web browsing, software engineering, cybersecurity, science and professional work.

Results and capabilities

According to figures published by OpenAI, the model scores 97.6% on FrontierMath Tier 4 (vs 83% for GPT-5.6 Sol), 99.9% on ARC-AGI-3, 100% on ExploitBench and 72.6% on OSWorld 2.0. OpenAI states that Astra completes computer-use tasks roughly 47% faster than its predecessor.

Availability and pricing

The model is rolling out in stages — first to selected organizations, then to ChatGPT Plus, Pro, Business and Enterprise subscribers, and via the OpenAI API, Microsoft Azure and AWS Bedrock. API pricing is USD 10 per 1M input tokens and USD 50 per 1M output tokens (Standard); a Fast mode offers up to 2x speed at 2x the price.

OpenAI has not disclosed the parameter count, context window size or the model's knowledge cutoff date.

Classification

LLMMultimodalReasoning modelTool-using model

Family: GPT

Access & deployment

APIHosted

Cloud

Weights: Closed

Key parameters

📥 Input: text, image

Technical specification

Modalities

⬇ Input

textimage

⬆ Output

textcode

Capabilities and applications

Native model capabilities

Reasoning

The model's ability to reason logically and solve complex problems.

Category: reasoning

Advanced reasoning

The ability to perform multi-step, structured reasoning: analysing problems, planning steps, and drawing conclusions from hypotheses. Reasoning-first models (e.g. GPT-5.1 Thinking) dedicate a portion of inference to chains of thought before responding.

Category: reasoning

Multi-step reasoning

Carrying out multi-step chains of reasoning across long, complex tasks.

Category: reasoning

Coding

Generating, analysing and modifying code in many programming languages. Covers writing functions, debugging, refactoring, code review, and creating tests. Measured by benchmarks such as HumanEval and SWE-bench.

Category: coding

Agentic coding

Multi-hour, multi-step programming tasks performed autonomously by the model: cloning a repository, running tests, iterating on fixes, integrating with CLI tools. Characteristic of Codex variants (GPT-5.1-Codex-Mini, Codex-Max).

Category: coding

Agentic capability

The model's ability to autonomously plan and execute multi-step tasks by sequentially using tools, maintaining context, and adapting to intermediate results.

Category: planning

Computer use

The model's ability to operate a computer interface by interpreting screenshots and generating actions such as clicks, typing, and navigating applications.

Category: planning

Web browsing

Ability of the model to autonomously search and browse web pages to retrieve up-to-date information.

Category: other

Image understanding

Analysing and interpreting the content of images.

Category: vision

Tool use

The model's ability to call external functions, APIs and tools during a conversation: calculator, search engine, code editor, database. The model decides when and how to use a tool and interprets its result.

Category: planning

Cybersecurity

The model ability to perform computer-security tasks: vulnerability analysis, proof-of-concept exploit generation, patching, and cybersecurity question answering.

Category: other

Vulnerability detection

The ability to identify and analyze security vulnerabilities in source code and software.

Category: coding

Planning

Forming and executing action plans for complex tasks.

Category: planning

Multi-step project execution

The ability to autonomously drive multi-hour, multi-step projects: decomposing the task, planning the sequence of actions, iteratively delivering results, and adjusting based on feedback. Key for enterprise and knowledge-work agents.

Category: planning

Benchmark results

5 benchmarks

EpochAI Frontier Math

accuracy · FrontierMath Tier 4

97.6%