Claude Fable 5 requires you to accept Anthropic’s data retention terms in your dashboard before you can use it. Your code is never used for training. Learn more in our docs.
used fable on low and asked it to fix a tiny bug, took like 3 minutes and used $1.72. idk when anthropic will just release a model thats fast for small tasks. composer 2.5 was better for what i needed and cost me like 13 cents
Lot’s of issues so far with it. I’ve had to contact support about the issue it has with parallel agents currently where the parallel agents almost always fail during streaming (and you’re still charged), and then can’t be resumed, etc.
I understand high demand but error handling should be improved on Cursor end for new flagship releases.
Also, are we notified when “Fable 5 runs with tight security guardrails. When a request trips one of those guardrails, Cursor routes it to Claude Opus automatically so your work continues without an error. You don’t need to retry or switch models yourself”. I don’t want to be routed to Opus without knowing. I’d rather just stop the chat at that point and use Opus.
Cursor team, I just wanted to say that Fable 5 is literally my most favorite AGI-like LLM model EVER, period. I will keep using Cursor as much as I can still, I love this IDE so much!! Thank you so much for adding this AI model for us to use!
Hello - I have enabled this but I still cannot chose Fable from the list of models in Cursor - what am I doing wrong? I have a Cursor Pro subscription.