GPT-6 Astra is OpenAI's most capable model, released on September 3, 2026 and built for long, end-to-end work: complex reasoning, coding, computer use, research and document creation.
OpenAI describes it as state of the art on computer use, browsing, software engineering, cybersecurity, science and professional work, and it is the first model the company has designated as reaching its “critical” cybersecurity threshold under its preparedness framework.
Availability note: this model is not yet served through AI/ML API. This page is published ahead of integration; pricing and the endpoint will be filled in once it goes live.
| Developer | OpenAI |
|---|---|
| Model ID | gpt-6-astra (OpenAI); AI/ML API id to be confirmed on release |
| Release Date | September 3, 2026 |
| Context Window | 1,050,000 tokens |
| Max Output Tokens | 128,000 tokens |
| Knowledge Cutoff | April 30, 2026 |
| Modalities | Image, Text → Text |
| Capabilities | function calling, reasoning (configurable effort), streaming, structured outputs, vision, web search, prompt caching |
Computer use
OpenAI reports the model can work through spreadsheets, fill in forms and navigate across web pages, often faster than a person doing the same task.
Very long context
A 1,050,000-token context window with up to 128,000 output tokens, so entire codebases or document sets fit in a single request.
Configurable reasoning
Depth of reasoning is controlled per request through the reasoning.effort parameter rather than by switching models.
Positioned as a generational step
OpenAI president Greg Brockman called it a “generational leap” that may eventually be seen as the arrival of AGI — a claim from the vendor, not an independent benchmark result.
Launch pricing on AI/ML API: Input: $13 / 1M tokens, Cached input: $1.3 / 1M tokens, Cache write: $16.25 / 1M tokens, Output: $65 / 1M tokens. The model is not yet callable via the API — pricing is published ahead of integration.
Requests over 272K input tokens are billed at 2x input and cache rates and 1.5x output for the whole request.
Agentic and computer-use workflows
Tasks that span many steps across real applications and web pages rather than a single prompt and answer.
Large-codebase software engineering
The context window and coding focus suit whole-repository work and long refactors.
Research and document production
Long-form synthesis over large source sets, with web search and file search available as tools.
Sources: https://openai.com/index/gpt-6-astra/ · https://developers.openai.com/api/docs/models/gpt-6-astra · https://www.axios.com/2026/09/03/openai-astra-gpt-6-agi-brockman
Measured on the AI/ML API gateway across live traffic. Figures are medians unless stated. Last updated 2026-10-12.
GPT-6 Astra is suited to complex reasoning, coding, computer use, research, and document creation.
It is documented as useful for software development, code generation, and repository analysis.
It is also documented for research, knowledge discovery, and information synthesis.
Another stated use case is computer-use and task execution across applications, including workflow automation with tools, agents, and connected systems.
import requests r = requests.post( "https://api.aimlapi.com/v1/chat/completions", headers={"Authorization": "Bearer " + AIMLAPI_KEY}, json={ "model": "openai/gpt-6-astra", "messages": [ { "role": "user", "content": "Hello!" } ] }, ) print(r.json())
const r = await fetch("https://api.aimlapi.com/v1/chat/completions", { method: "POST", headers: { Authorization: `Bearer ${process.env.AIMLAPI_KEY}`, "Content-Type": "application/json", }, body: JSON.stringify({ "model": "openai/gpt-6-astra", "messages": [ { "role": "user", "content": "Hello!" } ] }), }); console.log(await r.json());
curl -X POST https://api.aimlapi.com/v1/chat/completions \ -H "Authorization: Bearer $AIMLAPI_KEY" \ -H "Content-Type: application/json" \ -d '{"model":"openai/gpt-6-astra","messages":[{"role":"user","content":"Hello!"}]}'
OpenAI-compatible — swap the base URL and it works with your existing SDK.
| Type | Price |
|---|---|
| Input | |
| Output | |
| Cached input |
| Benchmark | Score | What it measures | Source | Retrieved |
|---|---|---|---|---|
| Intelligence | 52.8 | Composite score across standardised reasoning, knowledge and problem-solving evaluations, measured independently by Artificial Analysis | Source | September 12, 2026 |
| Coding | 76.9 | Composite score across standardised coding evaluations, measured independently by Artificial Analysis | Source | September 12, 2026 |
| Model | Input | Output | Context | Best for |
|---|---|---|---|---|
GPT-6 Astra This page | Agentic workflows and computer use | |||
| Agentic workflows and structured output | ||||
| Agentic workflows and structured output | ||||
| Agentic workflows and structured output |
GPT-6 Astra has a 1,050,000 tokens context window and can return up to 128,000 tokens.
GPT-6 Astra takes image, text as input and returns text.
Use openai/gpt-6-astra as the model id. Requests go to https://api.aimlapi.com/v1/chat/completions.
GPT-6 Astra became available on September 3, 2026.
GPT-6 Astra is priced at input $13 / 1M tokens, output $65 / 1M tokens, cached input $1.3 / 1M tokens.
Yes, GPT-6 Astra can stream responses as they are generated.
Yes, GPT-6 Astra accepts image input alongside text.
GPT-6 Astra was built by OpenAI.
No. It accepts image and text input but only outputs text.
Yes, it is available via the endpoint https://api.aimlapi.com/v1/chat/completions.
Yes, it supports function calling, parallel tool calls, and structured outputs.
