OpenAI Ultrafast Premium Speed Tier
OpenAI announced Ultrafast at DevDay 2026 on 29 September 2026, a premium speed tier for workloads where speed matters most. OpenAI says it offers up to 8x faster token generation (about 300 tokens per second) in Codex and up to 6x in the API.
Tool Overview
Features, steps and FAQ below
Features
- ✓ OpenAI says Ultrafast is its premium speed tier for workloads where speed matters most, announced at DevDay 2026
- ✓ OpenAI says it offers up to 8x faster token generation (about 300 tokens per second) in Codex and up to 6x in the API
- ✓ OpenAI says GPT-6 Astra Ultrafast is available in the API and in ChatGPT Work and Codex on Pro 500 and Enterprise plans
- ✓ OpenAI says GPT-6.1 Sol Ultrafast is coming soon
- ✓ OpenAI says Ultrafast is part of DevDay 2026's more than 20 major announcements, alongside Dots always-on agents, GPT-6.1 Sol, Private Intelligence and computer use in the Agents API
How to Use
- On a Pro 500 or Enterprise plan, use Ultrafast in ChatGPT Work and Codex
- Call Ultrafast through the API (OpenAI says GPT-6 Astra Ultrafast is available in the API)
- Watch for GPT-6.1 Sol Ultrafast to arrive
- Read the OpenAI DevDay 2026 recap to see Ultrafast and the other announcements from the event
FAQ
What is Ultrafast?
OpenAI says Ultrafast is a premium speed tier for workloads where speed matters most, announced at DevDay 2026.
How fast is it?
OpenAI says Ultrafast offers up to 8x faster token generation (about 300 tokens per second) in Codex and up to 6x in the API.
Which models and plans are available?
OpenAI says GPT-6 Astra Ultrafast is available in the API and in ChatGPT Work and Codex on Pro 500 and Enterprise plans, with GPT-6.1 Sol Ultrafast coming soon.
How do I get it?
OpenAI says you can call Ultrafast in the API, and Pro 500 and Enterprise plan users can use it in ChatGPT Work and Codex.
When did it launch?
OpenAI says Ultrafast was announced at DevDay 2026 on 29 September 2026, with GPT-6 Astra Ultrafast available that day in the API and in ChatGPT Work and Codex.