TL;DR — Fable 5 costs exactly 2× Opus 5 ($10/$50 vs $5/$25 per 1M tokens) — so the real question is which calls genuinely need Anthropic's top tier. Every number below is from our weekly-verified pricing tracker (verified 2026-08-26).
Pricing side by side (per 1M tokens, USD)
| Model | Provider | Input | Output | |-------|----------|-------|--------| | Claude Fable 5 | Anthropic | $10 | $50 | | Claude Opus 5 | Anthropic | $5 | $25 |
Verified 2026-08-26 against official rate cards — the tracker keeps the change log.
What a real month costs
At a standard workload of 10M input + 2M output tokens/month:
- Claude Fable 5: $200.00
- Claude Opus 5: $100.00
On this mix, Claude Opus 5 is the cheaper column (2.0× gap). Your ratio will differ — run your own mix through the LLM cost calculator.
Choose Claude Fable 5 if
you are escalating the hardest reasoning and agentic work, where a failed run costs more than the tokens. At low volume — the top 5% of traffic — paying 2× on those calls is cheap insurance.
Choose Claude Opus 5 if
you want a frontier default. The August 2026 repricing put Opus 5 at $5/$25 (a 67% cut from Opus 4's rates), which is what made it viable as a high-volume workhorse rather than a special-occasion model.
Routing decision card
- The ladder, not the choice: Opus 5 as frontier default, Fable 5 as the escalation-only rung above it
- Both sit behind Anthropic's ~90% cache discount and −50% batch API, which stack — cached, batched Opus 5 input lands near value-tier pricing
- Escalate by failure signal (retry, low confidence), not by guessing task difficulty up front
- The router answer: most teams shouldn't hard-pick one — ClawRouters routes each request to whichever side fits it, so the comparison becomes a per-call decision instead of a bet.
Per-model detail: Claude Opus 5 deep dive
Part of the LLM model comparison index. Prices move fast in 2026 — the tracker logs every change with dates.