From the source
Anthropic introduced Claude Fable 5.2 and Mythos 5.2, saying both handle complex software engineering tasks better while cutting API prices by 20% against the previous versions.
- The two share the same underlying model but differ in safeguards: Fable is broadly available, while direct access to Mythos is restricted to vetted cybersecurity and life sciences organizations to limit misuse.
- Both are priced at $8 per million input tokens and $40 per million output tokens, down from $10 and $50, and Anthropic says they generate output up to 30% faster than their predecessors.
- Anthropic reported that Fable 5.2 scored 49.1% on the FrontierCode 2.0 benchmark at its Xhigh effort setting, against 32.4% for Fable 5.1, and that at its High effort setting it matched Opus 5.5 at its highest setting at roughly half the cost per task.
- Mythos 5.2 identified 89.6% of vulnerabilities in Anthropic's internal CVE Rediscovery evaluation, up from 78.3% for Mythos 5.1, across 1,829 previously disclosed vulnerabilities.
- Anthropic recommends Opus 5.5 for most coding and knowledge work, with Fable and Mythos intended for more demanding assignments, and says customers eligible for zero data retention can keep those arrangements while others face 30-day retention for safety monitoring.