What the measurement shows
Anthropic's Fable 5.2 completed 41.3 percent of real freelance jobs at professional quality on an evaluation called the Remote Labor Index. This suggests the model has reached a notable level not just on test questions but on real work outputs.
Gap with rivals
The next-best systems on the same index were Opus 5.5 and GPT 6 Astra. Fable 5.2 reached more than double their score.
A big jump in a year
A year ago the top score on the same index was 2.5 percent, meaning the rate of completing real jobs has grown roughly sixteenfold in a year.
Weak spots
Results are not uniform across fields. Product design and animation remain weak spots for the models. That suggests the impact of automation on the workforce will advance at different speeds depending on the sector.
