01
What is GPT-6 Astra?
OpenAI released GPT-6 Astra on September 3, 2026 and describes it as its most capable model for difficult end-to-end work. The official API model ID is gpt-6-astra.
The model accepts text and image input and produces text output. OpenAI lists a 1,050,000-token context window, up to 128,000 output tokens and an April 30, 2026 knowledge cutoff.
02
Release date and availability
Astra launched with a limited enterprise-first rollout. OpenAI's current documentation says API access and availability through Plus, Pro, Business and Enterprise plans are coming in the following days rather than being universally available at launch.
03
GPT-6 Astra API pricing
OpenAI currently lists Standard text pricing per 1 million tokens at $10 input, $1 cached input, $12.50 cache writes and $50 output. Cache writes are billed at 1.25x the uncached input rate.
Prompts above 272K input tokens are priced at 2x input and cache rates and 1.5x output rates for the full request. Batch and Flex are priced at 50% of Standard, while Fast mode is 2x the applicable rate where supported.
04
What changed for agents and tool workflows
Astra adds async tool calling, allowing the model to continue reasoning or work on independent parts of a request while an application-side tool is still running. The application remains responsible for executing the tool and returning its result with the original call ID.
Mid-turn steering lets applications send new user instructions while Astra is already working over a WebSocket Responses API connection. OpenAI also documents configuration updates that can change reasoning effort during a conversation while preserving the cached prompt prefix.
05
Tools, rate limits and current limitations
Through the Responses API, OpenAI lists web search, file search, image generation, code interpreter, hosted shell, apply patch, Skills, computer use, MCP and tool search as supported. Function calling, streaming and Structured Outputs are also supported.
Fine-tuning is currently unsupported. The model page lists no Free-tier API access; paid rate limits vary by usage tier, from Tier 1 at 500 RPM and 500,000 TPM up to Tier 5 at 15,000 RPM and 40,000,000 TPM at publication time.
06
Who should use GPT-6 Astra?
Astra is most compelling when task completion matters more than lowest token price: complex software engineering, browser or computer workflows, research across many sources, large document sets and long-running agent tasks with multiple tools.
For routine extraction, classification or high-volume chat, lower-cost models may still be more economical. OpenAI says Astra can reduce estimated cost per successful task in some evaluations despite higher per-token prices, but that remains a vendor claim that should be tested on your own workload.
Sources
Primary and supporting sources
Facts were rechecked against the linked sources immediately before publication. Pricing, product availability and rollout status can change.