In the Responses API, set model to gpt-6.1-sol and service_tier to ultrafast. Per million tokens for short-context requests, Standard→Ultrafast rates are $2→$12 input, $0.10→$0.60 cached input, $2.50→$15 cache writes, and $10→$60 output. Requests above 272K input tokens use separate long-context rates; regional processing may add a 10% premium under the official price card. Default Ultrafast limits are 1M, 4M, and 40M tokens/minute for Build, Launch, and Grow respectively. Persistent WebSocket connections can reduce network overhead for multi-step agents. The six-times factor is a price multiplier, not a sixfold speed guarantee; no default migration of Standard requests or legacy-plan transition has been announced.