Model comparisons

Every comparison pulls its numbers from a single sourced dataset. Figures Anthropic has not officially published are flagged unofficial.

• Fable 5 vs Mythos 5: Same Model, Different Safeguards Fable 5 and Mythos 5 share one underlying model — the only real difference is safeguards. Here's what changes and which edition you'd actually use.
• Fable 5 vs Opus 4.8: Mythos-Class vs the GA Flagship Fable 5 sits a tier above Opus 4.8 but costs more and is currently suspended. Opus 4.8 is generally available. Here's how to choose between them.
• Mythos vs Opus: Tier and Access, Side by Side Mythos 5 is a tier above Opus 4.8, but it's Glasswing-only and currently suspended. Opus 4.8 is the open flagship. Here's the honest comparison.
• Fable 5 vs GPT-5.5: A Cross-Vendor Reality Check Comparing Anthropic's Fable 5 with OpenAI's GPT-5.5. We document Fable 5 specs precisely and point you to OpenAI for authoritative GPT-5.5 numbers.
• Fable 5 vs Gemini 3.5: Cross-Vendor, Honestly Scoped How Anthropic's Fable 5 stacks up against Google's Gemini 3.5. We give Fable 5 specs precisely and defer to Google for authoritative Gemini numbers.
• Fable 5 vs Sonnet 4.6: Frontier Power vs Everyday Value Fable 5 is Mythos-class and premium-priced; Sonnet 4.6 is the balanced, cheap workhorse. Here's when each one is the right tool for the job.
• Opus 4.8 vs Sonnet 4.6: The Practical GA Choice For most users the real decision is Opus 4.8 vs Sonnet 4.6. Both are generally available — here's how to pick on capability, cost, and workload.