Inside the Claude 5 family: when 2× pricing for Fable pays for itself.
claude-fable-5 · anthropic
claude-opus-5 · anthropic
Benchmark
| Workload Scenario | Claude Fable 5 | Claude Opus 5 | Price Delta |
|---|---|---|---|
| 1M input tokens (raw text) | $10.00 | $5.00 | +$5.00 |
| 1M output tokens (generation) | $50.00 | $25.00 | +$25.00 |
| 1M tokens · 70% input / 30% output mix | $22.00 | $11.00 | +$11.00 |
| Standard chat turn (4K in / 800 out, 50% cached) | $0.062 | $0.031 | +$0.031 |
| Monthly scale (10K requests / day) | $18,600.00 | $9,300.00 | +$9,300.00 |
Negative difference = Claude Fable 5 is cheaper. Positive = Claude Opus 5 is cheaper.
Scaling Curve
Total cost of a token volume at 70/30 input/output split (uncached). Log-log scale.
Verdict
Claude Opus 5 is cheaper on both input and output rates, so it costs less at every input/output mix. Price alone still isn't the whole decision: capability, latency and context limits (Claude Fable 5: 1M, Claude Opus 5: 1M) may justify the premium for your task.
FAQ
On input, Claude Opus 5 is cheaper ($5.00/M vs $10.00/M). On output, Claude Opus 5 is cheaper ($25.00/M vs $50.00/M). The same model is cheaper on both sides, so it wins at every mix.
A chat-style request (4,000 input + 800 output tokens, 50% cached) costs $0.062 on Claude Fable 5 and $0.031 on Claude Opus 5 — Claude Opus 5 is 2.0× more expensive for that workload.
Claude Fable 5 supports 1,000,000 tokens (128K max output); Claude Opus 5 supports 1,000,000 (128K max output). They are equal.