01 What happened
OpenAI’s documentation describes Ultrafast as the fastest service tier in the OpenAI API. It is currently available to all API users for GPT-6 Astra and GPT-6.1 Sol, subject to low rate limits.
02 Key details
- Default GPT-6 Astra limits are 500,000 tokens per minute for the Build tier, 1,000,000 for Launch, and 5,000,000 for Grow.
- In each response.create event, set model to gpt-6-astra and service_tier to ultrafast. OpenAI strongly recommends WebSockets for applications making frequent, rapid tool calls: without a persistent connection, network overhead can reduce latency gains.
- Ultrafast also supports HTTP requests through the SDK. OpenAI recommends using it when speed justifies the higher cost.
- Ultrafast mode for GPT-6.1 Sol supports US and EU data residency and global processing. GPT-6 Astra Ultrafast supports US data residency and global processing only.
- Organizations that work with an OpenAI account team can contact the team to request higher rate limits.
03 Why it matters
Developers can select OpenAI’s fastest API service tier for both GPT-6 Astra and GPT-6.1 Sol, considering higher cost, low rate limits, and regional processing restrictions that differ by model.
04 Who it matters to
Software developers.
Original source