The message announces re-opening the $200 Pro subscription to new buyers while changing how its included usage is calculated so that, by the poster’s math, subscribers will effectively spend half as much on API usage compared with the previous Pro $200 plan. Key specifics: the prior 5‑hour limit won’t return, subscribers are promised increasing work-per-dollar as models become more efficient and API prices fall, and recent cuts include GPT‑6 Sol and GPT‑6 Luna at 50% of their previous prices. The stated intent is to pass model efficiency gains to customers through lower API costs rather than mask value with inflated list prices and discounts. Additional subscription features that won’t consume usage are being added shortly, with fuller announcements coming the next day.
The rationale emphasizes long‑term value: subscriptions should deliver growing capability and lower effective unit cost, and over time pay‑as‑you‑go usage should approach parity with subscription value. Reaction in replies highlights friction points: users question how changes affect existing subscribers mid‑billing, compare value to competing offerings (e.g., Opus 5.5), and express frustration about company shifts and vendor choices. The post frames the changes as transparent groundwork ahead of a larger set of subscription improvements.
Summary generated by AI from the linked article. hn.today is not affiliated with Hacker News or Y Combinator.