Some people are not the sharpest tool in the shed
What you should be looking at is the high effort from Fable 5.2, which is already the best model, even when comparing against max effort models
You can wield frontier level intelligence for $9 or lower, if you tune it down to medium
You only ever need to use Fable 5.2 max effort for the most difficult and complex of tasks
What you should be looking at is the high effort from Fable 5.2, which is already the best model, even when comparing against max effort models
You can wield frontier level intelligence for $9 or lower, if you tune it down to medium
You only ever need to use Fable 5.2 max effort for the most difficult and complex of tasks
diminishing returns hit like a freight train.
claude fable 5.2 takes the intelligence crown at 65 (vs opus 5.5 at 59). but look at the damage:
- $26.54 per task (vs $10.04 on opus 5.5)
- 46.4 minutes per task (vs 25.1 mins)
you are paying +164% more money and waiting 85% longer for a 6 point benchmark bump.
we went from "ai will make intelligence free" to taking out a loan to resolve a single task.View quoted post on X
