GPT-6 Astra (Azure)

GPT-6 Astra surfaces inside the Microsoft Foundry ecosystem, as reflected by a Lambert Consulting briefing titled "GPT-6 Astra in Microsoft Foundry: Access and Data" published on the model's launch day. The model sits alongside other Azure-routed entries, making it available to enterprise customers who already build on Microsoft's AI stack for retrieval-augmented assistants, document workflows, and back-office automation. Its placement under the Foundry umbrella signals an intent to serve workloads that benefit from tight integration with Azure data services and governance tooling rather than as a standalone consumer-facing product.

The model's first high-profile deployment is the OpenAI Daybreak cybersecurity pilot, in which GPT-6 Astra's capabilities are being offered to MS-ISAC partner organizations supporting state and local cyber defenders as well as water-utility systems. This positions the model squarely in the critical-infrastructure protection space, where reasoning over threat intelligence, summarizing incident reports, and assisting defenders with triage workflows are practical strengths. The Daybreak framing suggests GPT-6 Astra is intended for teams that need an AI partner capable of sustained analytical work on operational security data, complementing human analysts rather than replacing them.

LLM Gatewayazure/gpt-6-astragpt-astra

Open provider page Provider docs

Quick Info

Provider: LLM Gateway
Model key: azure/gpt-6-astra
Release date: Sep 4, 2026
Last updated: Sep 4, 2026
Knowledge cutoff: 2026-04-30
Input modalities
Output modalities
Capabilities

Cost

Input token cost: $10.00
Output token cost: $50.00

Limits

Input tokens: 922,000 tokens
Output tokens: 128,000 tokens
Context window: 1,050,000 tokens

Transparent token rates

Compare GPT-6 Astra (Azure) pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about GPT-6 Astra (Azure)

LLM Gateway

CoverageBenchmark

A hands-on review reports that OpenAI shipped GPT-6 Astra on September 3, 2026 and put it in front of paying users two days later, making it the most expensive model the company has ever sold on the public API. The review confirms access on day two via an OpenAI-compatible endpoint running the gpt-6-astra model id, wit Pricing is documented at $10/M input, $50/M output, $1 cached input, $12.50 cache write, $5/$25 batch, 2x above 272K input tokens, and a Fast mode at 2x; the reviewer confirmed the headline rate against a live call, an 11-token prompt returning 5 tokens billed at $0.00036. The review characterizes the gains as real but

LLM Gateway

Coverage

Microsoft CEO Satya Nadella confirmed on September 4, 2026, that early customers are already using GPT-6 Astra on Azure, offering an early signal of how rapidly OpenAI's newest frontier model is moving from launch into enterprise environments. The update links to Microsoft's announcement that GPT-6 Astra began rolling The rollout highlights named enterprise users including Replit, whose CTO Luis Hector Chavez said the model opens up a new level of agentic capability beyond traditional code generation and can support active software creation, and Albertsons Companies, cited by Microsoft for its use of Azure OpenAI. The piece undersco

LLM Gateway

CoverageBenchmark

An independent benchmark explainer covers OpenAI's September 3, 2026 release of GPT-6 Astra, which the announcement describes as "the world's most intelligent and aligned model" and which Greg Brockman told reporters is "not unreasonable to feel that we are now in the AGI era." Under the hood it is OpenAI's largest tra On computer use, Astra posts 59.3 on Agents' Last Exam versus 53.6 for GPT-5.6 Sol, 72.6% on OSWorld 2.0 versus 65.7% for Sol (completed in about 40 minutes per task instead of 75, a 47% time reduction), and 92.7% on ScreenSpot-Pro against 76.9% for Sol; Mind2Web browser tasks run 1.9x faster than the current Sol setup

LLM Gateway

CoverageBenchmark

OpenAI's GPT-6 Astra launched on 2026-09-03 in a phased rollout: participants in OpenAI's cybersecurity access program first, then ChatGPT Plus, Pro, Business, and Enterprise tiers, followed by the API (model id gpt-6-astra), Azure, and AWS Bedrock. The model ships with a 1.05M-token context window, 128K max output, a Pricing is set at $10/M input and $50/M output, with cached input at $1.00/M and cache writes at $12.50/M; requests above 272K input tokens are billed at 2x input and 1.5x output for the entire request. On independent benchmarks, GPQA Diamond reaches 96.1% at max effort (Artificial Analysis, rank 1 on the AA board at i