Fable 5 vs Opus 5
Fable 5 is the previous Fable snapshot, still sold at $5/$25 — twice Opus 5. Anthropic rates it slower, and the cutoff is older. For new Fable work use 5.1. This comparison is for people who still land on the old slug.
Claude Fable 5 is still callable on llmrelay and still sold, but superseded. The comparison below is accurate; just know there is a newer generation if you are starting fresh.
Input / M (llmrelay): Anthropic list: $10 and $5.
Anthropic latency rating: Fable 5 is the only model Anthropic labels Slower. Adaptive thinking is always on and cannot be turned off, which is part of why.
Reliable knowledge cutoff: The more expensive model has the older cutoff. Worth knowing before assuming the flagship is strictly better.
Adaptive thinking: On Fable 5 you cannot disable it, so you pay for reasoning tokens on easy prompts too.
Prices are llmrelay's, at 50% of official list. Specs are the vendor's own published figures (Anthropic model docs), not our benchmarks. We do not publish scores we cannot source.
Use Fable 5 when
- +You have a genuinely hard problem and have already watched Opus 5 fail at it. That is the only honest trigger — capability ceilings show up as failures, not as a hunch.
- +The cost of a wrong answer dwarfs the token bill: migrations you only run once, architecture you will live with for years, analysis someone will act on.
- +Throughput does not matter and latency does not matter.
Use Opus 5 when
- +You want the current default. Anthropic's own docs say to start with Opus 5 for complex agentic coding and enterprise work, and reach for Fable 5.1 only when Opus evals fall short.
- +You need the fresher knowledge cutoff — May 2026 against January.
- +You are paying per task at any real volume. At 2× the rate and always-on thinking, Fable 5 bills escalate faster than the price ratio suggests.
The honest answer
Opus 5 is the default. If you escalate, escalate to Fable 5.1, not this snapshot. Fable 5 stays on the key for pinned output you already validated — at twice the Opus rate, with an older cutoff, and Anthropic still rates it slower.
What this costs you per month
At 50M input and 5M output tokens a month — a realistic heavy agent workload.
Struck-through column is the vendor's list price for the same traffic. Adjust the numbers on the calculator.
One key covers both
Switching between them is a model-id change, not a new account. Prepaid, no subscription, $10 minimum — enough to run the same prompt through both and settle the argument with your own output.
Next steps on this decision
Per-model detail for both sides, and the adjacent comparisons people usually read alongside this one.
- Claude Fable 5 pricing and specs$5.00/M in, $25.00/M out — half list. Previous Fable. Same $10/$50 list, still served. Use 5.1 for new work.
- Claude Opus 5 pricing and specs$2.50/M in, $12.50/M out — half list. Current flagship Opus. A generation past 4.8, same list price.
- Opus 5 vs Sonnet 5 for codingBoth are current-generation Claude models with a 1M context window.
- Opus vs Sonnet for coding (measured)Measured on our own billing: $0.0152 against $0.0047 for one identical task.
- GPT-6 Astra vs Opus 5 — is 2× worth it?Astra is OpenAI's current flagship.
- Cost calculatorPut your own monthly token volume in and see the bill against vendor list price.