GPT-6 Astra: prompts over 272K tokens double the rate on the entire request

OpenAI shipped GPT-6 Astra on September 3, 2026, with a ~1.05M-token context window and $10/$50 per million input/output tokens. Requests whose input prompt exceeds 272,000 tokens are billed at 2× input and cache rates and 1.5× output rates on the whole request, not just the overflow.

• GPT-6 Astra offers a ~1.05M-token context at $10/$50 per 1M input/output tokens
• Prompts over 272,000 tokens: 2× input and cache, 1.5× output on the entire request
• A 275K-token request costs $6.25 vs $3.20 at 270K — a $3.05 jump
• The surcharge isn't flagged in the API response; teams must check usage.input_tokens