GPT-6 Astra is built for difficult work that takes more than one step

GPT-6 Astra is OpenAI's most capable model for complex, end-to-end work. It is designed for jobs such as working through a codebase, researching across the web, using software and producing finished documents—not just answering a single question.

Astra can keep working while an application runs a tool, and a user can send new instructions before the task finishes. It supports text and image input, a 1.05-million-token context window and up to 128,000 output tokens. API pricing is $10 per million input tokens, $1 for cached input and $50 for output.

That makes Astra a premium option. It makes the most sense when a cheaper model needs repeated attempts or a lot of human correction. For routine chat and high-volume processing, Sol, Terra or Luna may still deliver better value.

What this means

• Astra targets long, tool-heavy coding and research tasks.
• Its API price is higher than the GPT-5.6 family.
• Judge it by successful task cost, including retries and review time.