TL;DR
Anthropic’s Claude Fable 5 is now the company’s most expensive active public model, priced at $10 per million input tokens and $50 per million output tokens. Independent benchmark data cited by Thorsten Meyer AI shows a 5.7% Intelligence Index gain over Opus 4.8, while public productivity evidence remains limited to a single unaudited customer story.
Anthropic’s Claude Fable 5 has become the company’s most expensive active public model, priced at $10 per million input tokens and $50 per million output tokens, while independent benchmark data cited by Thorsten Meyer AI shows a 5.7% Intelligence Index gain over Opus 4.8, a model listed at half the price.
The pricing details are described by Thorsten Meyer AI as confirmed by Anthropic’s launch materials, platform pricing documentation and product pages, with secondary corroboration from Forbes, TechCrunch, Vellum, Finout and Artificial Analysis. The article says Fable 5 costs 2.0 times more than Claude Opus 4.8 and about 3.3 times more than Sonnet 4.6.
The benchmark comparison is less favorable for buyers focused on aggregate price performance. According to Artificial Analysis data cited by Thorsten Meyer AI, Fable 5 scored 64.9 on the Intelligence Index, compared with 61.4 for Opus 4.8. On GDPval-AA, described as a knowledge-work Elo benchmark, Fable 5 scored 1,932 versus 1,890 for Opus 4.8.
The source material also says a full Intelligence Index run costs about $9,940 on Fable 5, compared with roughly $4,970 on Opus 4.8. It notes that effective costs can fall through prompt caching, batch pricing and usage mix, with a cited blended rate of $7.70 per million tokens under a 7:2:1 cache-hit-to-input-to-output ratio.
The development matters because enterprise AI buyers often justify premium models through expected gains in staff productivity, automation quality or higher-value work completion. In this case, the source material says the pricing is well documented, while the public evidence for broad productivity gains is much thinner.
For budget owners, the key issue is not whether Fable 5 is stronger than Opus 4.8. The cited benchmarks say it is. The question is whether the incremental performance is large enough on a buyer’s own tasks to justify a 2.0 times price increase over Opus 4.8.
The answer may vary by workload. Thorsten Meyer AI reports that Fable 5 sets records in 5 of 10 sub-benchmarks, meaning the premium could be more defensible for specialized tasks where the model’s gains are concentrated. For average workloads, the public numbers point to a narrower case.
AI language model API access
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
How Fable Compares Inside Claude
Fable 5 sits above Claude Opus 4.8 and Claude Sonnet 4.6 in Anthropic’s cited active lineup pricing. Thorsten Meyer AI says Opus 4.8 is priced at $5 per million input tokens and $25 per million output tokens, while Sonnet 4.6 was around $3 input and $15 output per million tokens.
The same source says Fable 5 and Mythos 5 use the same underlying weights and carry the same pricing, differing in safeguards. That means reports that group the two models together on price are described as accurate, rather than a category error.
Anthropic’s earlier Opus 4 and 4.1 models are described as the only Claude models priced higher, at $15 input and $75 output per million tokens, but those are not identified as the current active comparison point in the source material.
“The economics are fully documented. The productivity story is not.”
— Thorsten Meyer AI
enterprise AI model subscription
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Productivity Claims Lack Baselines
The main unresolved issue is whether Fable 5 produces enough real-world productivity gain to justify its top-line pricing across typical enterprise deployments. According to Thorsten Meyer AI, there are no controlled human-baseline productivity studies in the public evidence reviewed.
The strongest public productivity example cited is a Stripe migration story from Anthropic’s launch post, involving a 50-million-line Ruby codebase. The claim is attributed to Anthropic and Stripe, but the source material describes it as unaudited and not an independently established productivity benchmark.
It is also unclear how well the benchmark advantage maps to actual spending decisions. A 42-point Elo lead on GDPval-AA may matter for some tasks, but the public data does not yet show how often that advantage converts into measurable savings, faster delivery or fewer errors in production settings.
AI model benchmarking tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Buyers Turn To Workload Testing
The next test for Fable 5 is likely to happen inside enterprise evaluations rather than launch pages. Buyers weighing the model against Opus 4.8, Sonnet-class alternatives or competing frontier systems will need task-level trials that measure cost per accepted output, failure rates and human time saved.
More public evidence may come from customer case studies, independent benchmark updates and vendor documentation. Until then, the confirmed record shows a clear pricing premium, a measurable benchmark gain and an open question over whether that gain pays for itself at scale.
AI model performance evaluation
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
What happened with Claude Fable 5 pricing?
Claude Fable 5 is listed at $10 per million input tokens and $50 per million output tokens, according to pricing cited by Thorsten Meyer AI. That is twice the price of Claude Opus 4.8.
How much better is Fable 5 than Opus 4.8 in the cited benchmarks?
Artificial Analysis data cited by Thorsten Meyer AI shows Fable 5 at 64.9 on the Intelligence Index and Opus 4.8 at 61.4. That is a cited 5.7% aggregate gain.
Does the public evidence prove Fable 5 makes workers more productive?
No controlled public human-baseline study is cited in the source material. The main example is an Anthropic customer story involving Stripe, which Thorsten Meyer AI describes as unaudited.
Could Fable 5 still be worth the higher cost?
Yes, for some workloads. The source material says Fable 5 leads in 5 of 10 sub-benchmarks, so the premium may be easier to justify for tasks where its gains are concentrated.
What should buyers watch next?
Buyers should look for task-level evaluations, audited productivity data and cost-per-output comparisons against Opus 4.8 and other models before treating broad productivity claims as established.
Source: Thorsten Meyer AI
