🔍 Read the full analysis: A 5.7% AI Intelligence Lift Raises The Price Twofold on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Anthropic’s Claude Fable 5 is priced at $10 per million input tokens and $50 per million output tokens, twice the listed rates for Opus 4.8. Artificial Analysis reports a 5.7% lead on its Intelligence Index, while the source review finds no controlled human-baseline productivity study and identifies one unaudited customer testimonial.
Anthropic’s Claude Fable 5 costs twice as much as Opus 4.8 at list price, while third-party benchmark data cited by ThorstenMeyerAI.com shows a 5.7% lead on the Intelligence Index and the site’s review found no controlled study demonstrating its productivity gains against a human-work baseline. The figures put a documented price premium alongside a much thinner public record for the workplace benefits that could justify it.
Anthropic lists Fable 5 at $10 per million input tokens and $50 per million output tokens. Opus 4.8 is listed at $5 and $25, respectively; Sonnet 4.6 is listed at approximately $3 and $15. Those rates make Fable 5 the most expensive model in Anthropic’s active lineup, according to the source report. The report says the pricing is stated in Anthropic materials and corroborated by several secondary sources.
Artificial Analysis benchmark figures cited in the report give Fable 5 an Intelligence Index score of 64.9, compared with 61.4 for Opus 4.8. The same comparison puts Fable 5 at 1,932 on GDPval-AA, a knowledge-work Elo benchmark, versus 1,890 for Opus 4.8. The report calculates those differences as 5.7% on the Index and 2.2% on GDPval-AA. These are benchmark results, not measurements of how much faster or more productive employees are in ordinary workplace use.
The source review says the full Intelligence Index run costs about $9,940 for Fable 5, compared with about $4,970 for Opus 4.8. It also describes lower-cost options that can affect actual spending: prompt-cache reads are priced at $1 per million tokens, and batch processing at $5 per million input tokens and $25 per million output tokens. A blended rate of $7.70 per million tokens is presented for a particular 7:2:1 cache-hit-to-input-to-output mix; that figure depends on the stated usage pattern and is not a universal rate.
The numbers matter to companies choosing models for substantial workloads because token prices recur with use. A model that scores higher may still cost more per completed task, while a lower-priced model may be sufficient for routine requests. The relevant comparison is not only the sticker price or a general benchmark score, but the cost and quality on the buyer’s own tasks.
The report’s central distinction is between measured benchmark performance and demonstrated productivity. Artificial Analysis provides comparative test results; the source review says the public productivity case rests on a single customer story rather than a controlled, independently audited study. That leaves budget owners without a verified public estimate of how much time or labor Fable 5 saves across typical enterprise work.
There may be workloads where the premium is worthwhile. The report says Fable 5 set records in five of ten sub-benchmarks, suggesting its relative strengths are not evenly distributed across tasks. That claim does not establish a gain for any particular company, but it makes task-specific testing more informative than treating the overall 5.7% Index difference as a universal productivity estimate.
AI language model pricing comparison
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Launch Claims and Available Evidence
Fable 5 was announced on June 9, 2026, and the source report places its general availability around July 1. ThorstenMeyerAI.com says the model’s price is confirmed in Anthropic’s launch announcement, platform pricing documentation and product page. It also notes that Fable 5 and Mythos 5 use the same underlying weights and have identical pricing, with safeguards distinguishing the products.
The report identifies Anthropic’s launch post as the source of the main productivity example. Anthropic attributed to Stripe a migration across a 50-million-line Ruby codebase that Fable 5 completed in a day, work the customer said would otherwise have taken a team more than two months by hand. The source review confirms that the claim appeared in the launch material; it does not independently verify the work, its comparison method or whether the result generalizes.
That distinction is important when describing the evidence. Repetition of a customer example by other publications does not by itself amount to an independent audit or controlled productivity study. ThorstenMeyerAI.com says its review examined 25 claims and rejected stronger descriptions of the Stripe account as an independently established human-work baseline.
“twice as much for 5.7 percent more performance”
— The Decoder, as quoted by ThorstenMeyerAI.com
enterprise AI model with best benchmark scores
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Productivity Claims Still Unverified
The public evidence cited in the source report does not establish how Fable 5 changes productivity across enterprise knowledge work. It reports zero controlled human-baseline studies and one unaudited customer testimonial. No methodology, independent audit or replication is provided for the Stripe example in the material supplied.
It is also unclear how the benchmark gap translates to particular company workloads. The report says Fable 5 leads on some sub-benchmarks, but does not provide enough detail here to determine which tasks any specific buyer performs or whether those gains offset the higher inference cost. The 5.7% Index figure should not be read as a 5.7% increase in worker output.
Some pricing comparisons also have limits. The source describes the Fable 5 comparison with GPT-5.5 as resting on a single secondary source and treats it as directional. Actual bills will depend on input and output volume, caching, batch use and any data-residency settings; the listed rates alone do not establish a buyer’s total cost.
AI model productivity benchmarking tools
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Measure Costs on Real Tasks
The next useful evidence would be task-specific evaluations that compare Fable 5 with lower-priced models on the same work, using disclosed methods and recording both quality and time. Independent replication of the Stripe account, or controlled studies against a defined human baseline, could clarify whether the anecdote represents a repeatable productivity gain.
Until such results are available in the source record, organizations weighing the model can compare actual token usage and output quality on representative tasks rather than assume benchmark improvements translate directly into labor savings. The source material does not identify a scheduled independent study or another confirmed productivity-data release, so the timing of further evidence remains unknown.
large language model input output tokens
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Key Questions
How much does Claude Fable 5 cost?
Anthropic’s listed rates, as reported by ThorstenMeyerAI.com, are $10 per million input tokens and $50 per million output tokens. Cache reads and batch processing have separate lower rates.
How does its price compare with Opus 4.8?
The source report lists Opus 4.8 at $5 per million input tokens and $25 per million output tokens. On those list rates, Fable 5 costs twice as much.
What does the 5.7% figure measure?
It is the reported difference between Fable 5’s 64.9 and Opus 4.8’s 61.4 on Artificial Analysis’s Intelligence Index. It is a benchmark comparison, not a measured increase in employee productivity.
Is there proof that Fable 5 improves workplace productivity?
The source review reports no controlled human-baseline study and identifies one Stripe-attributed customer testimonial in Anthropic’s launch post. It says the account has no disclosed audit or independent replication.
Could Fable 5 still be worth the higher price?
That depends on the buyer’s tasks, usage and required output quality. The report says Fable 5 led on some sub-benchmarks, but companies would need to test representative workloads and compare quality and cost per completed task to establish whether the premium pays off for them.
Source: ThorstenMeyerAI.com
Halloween Picks
halloween
As an affiliate, we earn on qualifying purchases.
