New top Claude models are out; Anthropic says start with Opus 5.5
On Nov. 9, Anthropic released Claude Fable 5.2, which it calls its most capable model for demanding technical work and says is available to all users. It also released Mythos 5.2, the same model with looser safeguards, open only to approved cybersecurity and biomedical organizations.
Anthropic itself tells most people to start with Claude Opus 5.5 for everyday coding and knowledge work. It suggests moving up to Fable only for unusually hard, open-ended projects, or down to the lighter Sonnet or Haiku models for simpler jobs. If you're learning to code or building your first app with Claude, the newest name is not automatically the right pick. The performance claims and glowing tester quotes come from Anthropic and its partners, not independent testing.
Worth knowing mainly so you don't chase the flashiest model. If you're starting out or using Claude for everyday work, stick with Opus 5.5 and skip Fable until you hit a problem Opus can't handle.
Your moveOpen the model picker in Claude and check which model you're using. Start with Opus 5.5, and switch to Fable 5.2 only when a project gets stuck.
Read the source