A temporary usage offer deserves a comparison with an objective before it becomes an excuse to open more agents without knowing why. On June 30, 2026, Anthropic announced Fable 5's return for July 1, accompanied by a temporary policy of included usage for subscriptions. The time limit belongs in the decision from the start.
On June 17, I had already asked to resume agents stopped by Claude usage limits. The task needed to continue and access was getting in the way. That is why I pay attention to usage conditions alongside answer quality. Available capacity changes what can be tried and how long the work can keep moving.
The announced window interests me as a chance to answer a specific question about usefulness. I want to know which task the model helps with and what review its delivery requires. Increasing volume merely because the offer allows it can produce plenty of conversation without improving that answer. Using every bit of capacity takes work too; I want something useful to show for it.
The cost of an experiment and the cost of maintaining a routine need separate treatment. Favorable conditions can justify a test I would otherwise postpone. When the offer ends, the workflow has a different bill. A decision about continued use requires examining that bill again, without turning the memory of a promotion into a permanent product condition.
I also want to know how a task that grows beyond expectations leaves its state when it must stop. That follows directly from my June resumption request. Even during an offer, execution needs to preserve enough information to continue. Commercial conditions change the starting point; they do not explain how interrupted work can be recovered.
To use this kind of window, I will keep the tasks that best reveal the model's usefulness, alongside their requests and the checks for their answers. That makes evaluation repeatable under normal conditions. The comparison remains useful even if I drop the choice: I can still explain which work it helped and what keeping access required.
My interest in Fable's return is using the window to learn something that outlasts it. The offer can justify an experiment, with its deadline recorded. Continued adoption comes after recalculating the cost and checking delivery under the following conditions. I want to leave the promotion with a decision I can explain, without depending on a memory that it seemed cheap at the time.