GPT-6.1 Sol Ultrafast opens to every API user at $12 / $60, six times standard Sol

OpenAI's Ultrafast tier now covers GPT-6.1 Sol for every API user. OpenAI's API docs, read on October 8, say Ultrafast mode is "broadly available" for GPT-6 Astra and GPT-6.1 Sol and "available to all API users", with preview access for GPT-5.6 Sol. It is a service tier, not a new model: you call gpt-6.1-sol with service_tier set to "ultrafast". OpenAI Devs announced the rollout the same day, calling it near-Astra intelligence at up to 8x the speed of Sol Standard.

THE PRICE

OpenAI's pricing page, per million tokens: - GPT-6.1 Sol Ultrafast: $12 input, $0.60 cached input, $15 cache writes, $60 output; $24 / $1.20 / $30 / $90 for long context - GPT-6.1 Sol Standard: $2 / $0.10 / $2.50 / $10, so Ultrafast costs six times as much - The Fast tier in between: $4 / $0.20 / $5 / $20 - GPT-6 Astra Ultrafast: $60 input and $300 output

HOW TO USE IT

Set service_tier to "ultrafast" on gpt-6.1-sol. OpenAI recommends WebSockets for agent loops that make many tool calls, because per-request connection overhead eats into the speed gain. Ultrafast has its own rate limits, separate from Standard and Fast.

IN CODEX AND CHATGPT WORK

Not for every plan there. OpenAI Devs says access in Codex and ChatGPT Work is on Pro 500, eligible usage-based Enterprise and credit-based Edu plans, and Enterprise admins must turn it on. Several X accounts read that as "Pro 500 only" across the board; the API itself is open to all users, per OpenAI's docs. Ultrafast for GPT-6.1 Sol runs in all supported regions, with US and EU data residency.

WHAT WE DO NOT KNOW

How fast it really is outside OpenAI's own "up to 8x": no independent tokens-per-second measurement has been published yet.