OpenAI launches "Ultrafast" mode for GPT-6.1 Sol: up to 8x faster, with restricted access in ChatGPT Work and Codex
Redacción IvokaOct 10, 9:00 AM 3 min
OpenAI is rolling out a faster mode for GPT-6.1 Sol across its API, Codex and ChatGPT Work. In ChatGPT Work and Codex, launch access covers $500 Pro plans and eligible Enterprise and Edu accounts.
OpenAI began rolling out "Ultrafast," a speed mode for GPT-6.1 Sol. According to OpenAI, it delivers near-Astra-tier intelligence at up to 8x the speed of Sol Standard.
Where it is available
It is rolling out in the API, Codex and ChatGPT Work. In the API, it is used with the gpt-6.1-sol model in the Responses API by setting service_tier to ultrafast. In Codex and ChatGPT Work, launch access covers $500 Pro plans, eligible usage-based Enterprise accounts (an admin must enable it) and eligible credit-based Edu accounts. OpenAI says Ultrafast is available in all supported regions, including data residency support in the US and EU. The rollout is ongoing, so it does not reach every plan equally.
What it costs
In the API, Ultrafast costs $12 per million input tokens and $60 per million output tokens. Standard mode for the same model costs $2 and $10, so Ultrafast is about six times more expensive. In Codex and ChatGPT Work, usage draws from plan credits under documented rates.
What it means for you
If you are a creator or run a small business and use ChatGPT on a regular plan, this does not change your day to day: in ChatGPT Work and Codex, launch access is limited to the plans mentioned above. It is more relevant news for teams building agents or automations that need answers with less waiting, and OpenAI positions it for work where speed matters, such as debugging an outage, agents navigating apps and live experiences. There is no independent verification yet that quality matches Astra on real workloads.