Does the Fable 5.2 flop indicate that Anthropic models have reached their max potential?
retained text
Fable is reportedly 10x the size of Opus in terms of parameters, but after months of training and even with its huge size, it seems only comparable to Opus 5.5, and even worse in some cases. Has anthropic peaked?
document fetched 2026-11-12 00:43 · canonical https://reddit.com/r/ClaudeCode/comments/1xtbyg9/does_the_fable_52_flop_indicate_that_anthropic