Side-by-side comparison of pricing, 12 benchmarks, and generation speed.
| Metric | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) | GPT-4 |
|---|---|---|
| Input ($/M tokens) | $12.5 | $30 |
| Output ($/M tokens) | $50 | $60 |
Data from Artificial Analysis API — 12 benchmarks
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) is cheaper overall. Its blended price (3:1 input/output ratio) is $21.88/M tokens vs $37.50/M for GPT-4.
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) wins 7 out of 12 benchmarks compared to 0 for GPT-4. See the detailed benchmark chart above for per-category results.
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) generates tokens faster at 61 tok/s vs 44 tok/s. However, GPT-4 has lower time-to-first-token (1.15s vs 90.38s).
Choose based on your priorities: Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) for lower cost, Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) for stronger benchmark performance, and Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) for faster generation. For latency-sensitive apps, check the TTFT comparison above.