Anthropic lowers the price of its demanding technical models
More capable technical AI becomes more useful when teams can afford to use it repeatedly.

More capable technical AI becomes more useful when teams can afford to use it repeatedly.
What happened
Anthropic released Claude Fable 5.2 and Mythos 5.2, priced at $8 per million input tokens and $40 per million output tokens. Fable includes stronger misuse restrictions; Mythos requires organisational verification. Claimed capability improvements come from the company.
Why it matters
The models target demanding work such as coding and reasoning. Lower prices could make more complex workflows economical, but teams still need to assess results on their own tasks.
The bigger picture
Claims about scientific discoveries and increasingly autonomous research require independent scrutiny. A product release should not be treated as proof that those wider questions are settled.
