OpenAI sets GPT-6.1 Sol Ultrafast API pricing

OpenAI’s stated API rates for GPT-6.1 Sol Ultrafast are $12 per million input tokens and $60 per million output tokens.[1] Those prices are six times the standard rates for GPT-6.1 Sol, which OpenAI lists as $2 per million input tokens and $10 per million output tokens.[3] The difference applies to both input and output tokens, so teams estimating costs should assess each separately rather than relying on one blended rate. The quoted pricing post was timestamped Oct. 8, 2026, at 6:26 PM.[1]

OpenAI listed Ultrafast mode for GPT-6.1 Sol as generally available on Oct. 8, 2026, and says it can be used through the Responses API.[2] All API users can access the mode, subject to rate limits.[2] OpenAI also says the API supports global processing and US and EU data residency for Ultrafast.[2] Separately, OpenAI said the mode was rolling out in Codex and ChatGPT Work; access there is available on Pro 500, eligible usage-based Enterprise and credit-based Edu plans, with Enterprise administrators required to enable it.[1]

OpenAI says Ultrafast reduces the time between generated output tokens.[2] The company says it can deliver speeds of up to eight times those of Sol Standard.[1] OpenAI identified outage debugging, agents navigating apps and live experiences where every second counts as potential use cases.[1] The speed claim is a maximum comparison with Sol Standard; the cited information does not specify a typical speed or promise that every workload will reach that level.[1] Teams weighing the higher API rates can therefore compare the value of faster output against their own task requirements and token use.