Claude Fable 5.2 and Mythos 5.2
Points and comments are a snapshot, not live.
Anthropic releases Claude Fable 5.2 and Mythos 5.2 with faster, cheaper coding and reasoning.
Fable 5.2 and Mythos 5.2 show improvements in coding, reasoning, and vision over the previous version. They generate outputs up to 30% faster and cost $8 per million input tokens, $40 per million output tokens (or $16/$80 for fast mode). Benchmarks show gains over Opus 5.5 on agentic coding, scientific research, and cybersecurity. In testing, the models designed three working electronic devices, solved two open black-hole physics problems, formalized Fermat's Last Theorem in 30 hours with 1,000 agents, and demonstrated improved robotics control. Mythos offers reduced safety restrictions for cybersecurity and biomedical research applicants.
What commenters are saying
The top comments focus on a duplicate-post confusion over timestamps, with users explaining HN's re-up mechanism. No substantive discussion of the article's content appears in the comments.