Claude Fable 5.5 (Leaked Checkpoint) vs Claude Sonnet 5.5
A direct, empirical evaluation of Claude Fable 5.5 (Leaked Checkpoint) (Anthropic) against Claude Sonnet 5.5 (Anthropic).
Claude Fable 5.5 (Leaked Checkpoint)
Developed by Anthropic · Proprietary Frontier Dynamic Reasoner parameters
Terminal-Bench 4.0 autonomous shell execution (78.5% target), recursive self-critique, and 1.5M-token repository refactoring.
Claude Sonnet 5.5
Developed by Anthropic · Proprietary Frontier Dynamic Hybrid Reasoner parameters
Terminal-Bench 4.0 execution (70.6%), SWE-bench Verified coding (82.4%), GitHub Copilot multi-file edits, and low-latency autonomous agentic coding at $2/$10 pricing.
| Metric | Claude Fable 5.5 (Leaked Checkpoint) | Claude Sonnet 5.5 |
|---|---|---|
| SWE-bench Verified | 84.2% | 82.4% |
| MMLU-Pro | 94.6% | 88.4% |
| MATH-500 | 98.4% | 95.2% |
| Context Window | 1,500,000 tokens | 1,000,000 tokens |
| Input Pricing (per 1M) | $3.50 | $2.00 |
| Output Pricing (per 1M) | $17.50 | $10.00 |