🔍 Read the full analysis: Claude Fable 5.2’S Per-Task Price Tag At Each Effort Level on ThorstenMeyerAI.com
Get the latest gadgets delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Artificial Analysis data accessed Nov. 10 puts Claude Fable 5.2’s average cost per Intelligence Index task at $3.10 to $26.54 across five effort settings. Its score rises from 54 at Low to 65 at Max, while the High setting scores 62 for $9.08 per task. These are benchmark-specific API cost estimates, not a fixed charge for every real-world job.
Anthropic’s Claude Fable 5.2 costs an average of $3.10 per task at Low effort and $26.54 at Max on Artificial Analysis’s Intelligence Index, according to benchmark data accessed Nov. 10. The reported score rises from 54 to 65 across those settings, giving businesses a measure of the price of additional model effort on this particular test.
Artificial Analysis measured five settings—Low, Medium, High, Xhigh and Max—across nine evaluations covering agentic coding, business documents, science and knowledge. Its per-task figure is the average API cost for one task in that index, not a universal quote for any prompt or workplace workflow. The reported scores and costs are: Low, 54 and $3.10; Medium, 56 and $3.80; High, 62 and $9.08; Xhigh, 62 and $14.83; and Max, 65 and $26.54.
The results show a marked step between Medium and High: the score increases by six points for $5.28 more per task. Xhigh costs $14.83 while retaining the same rounded score as High. Max adds three points over High but costs nearly three times as much. Artificial Analysis reports average task times of 6.5 minutes at Low and 46.4 minutes at Max; its figures also show output-token use across the full index rising from 34.9 million to 275 million.
The reported token rates are the same across effort settings: $8 per million input tokens and $40 per million output tokens. The different estimated bills reflect how much the model uses, including output tokens spent reasoning. Anthropic says Fable 5.2 can produce answers up to 30% faster than its 5.1 predecessor. An optional fast mode is priced at $16 per million input tokens and $80 per million output tokens, with Anthropic saying it generates tokens up to 2.5 times faster.
Claude Fable 5.2’s per-task price tag at each effort level
Artificial Analysis benchmark data accessed November 10, 2026 shows how Fable 5.2’s score, cost and task time change across five effort settings. These are estimates for one benchmark, not fixed prices for real-world jobs.
Cost and score by effort
Average API cost per Intelligence Index task and reported score across five settings.
| Effort | Cost per task | Score |
|---|---|---|
| Low | $3.10 | 54 |
| Medium | $3.80 | 56 |
| High | $9.08 | 62 |
| Xhigh | $14.83 | 62 |
| Max | $26.54 | 65 |
Where the trade-off shifts
The benchmark shows different returns as effort and estimated cost rise.
The score rises six points, from 56 to 62, for $5.28 more per task.
The rounded score stays at 62 while estimated cost increases from $9.08 to $14.83.
Max adds three score points, costs $26.54 per task and takes longer on average.
What drives the bill
The benchmark estimates reflect token use, not a different standard token rate for each setting.
- $8 per million input tokens across effort settings
- $40 per million output tokens across effort settings
- Output-token use rises from 34.9 million at Low to 275 million at Max across the index
- Average task time rises from 6.5 minutes at Low to 46.4 minutes at Max
Make the benchmark useful
Use the published figures as a comparison point, then test representative work before choosing a setting.
- Track token use, elapsed time and output quality on your own tasks.
- Compare effort settings against your current model and budget.
- Check the benchmark’s scope, access date and pricing assumptions; results may change.
The Cost of More Model Effort
The figures turn an abstract model setting into a cost-and-latency decision. A team running many tasks could pay substantially more by selecting Max, while the benchmark shows a smaller score gain between High and Max than between Medium and High. That makes effort level relevant to API budgeting and service design, especially where jobs are frequent or people are waiting for a result.
The benchmark does not establish that High is the right choice for every business task. Artificial Analysis’s score aggregates a defined set of evaluations, and the score differences do not say how much value a particular company would get from a more expensive answer. Max also takes longer on average in this test. The results are best read as a comparison under one benchmark setup, not a guarantee of quality, cost or timing in production.
Products Worth Considering
LG UltraGear 32GX850A-B 32-inch 4K UHD OLED Gaming Monitor
LG UltraGear 27-inch QHD OLED Gaming Monitor, 240Hz
Apple MacBook Air 13.6-inch Laptop with M5 Chip, 16GB Memory, 512GB SSD, Midnight
As an affiliate, we earn on qualifying purchases.
Fable 5.2 and Its Effort Settings
Anthropic announced Claude Fable 5.2 on Nov. 9, 2026, alongside Claude Mythos 5.2, and said the models replace their 5.1 versions. Medium is the default effort setting in Claude apps and Claude Code, while users can select Low, High, Xhigh or Max. Anthropic describes Fable as available to users with safeguards for high-risk cybersecurity work and most research biology tasks. Mythos is described as a version for organizations accepted into Anthropic’s Cyber or Life Sciences Verification Programs.
Artificial Analysis ranks Fable 5.2 at No. 1 of 74 models at Max, with a score of 65. The same model occupies the next two positions in the reported ranking at Xhigh and High, both with rounded scores of 62. The ranking is a snapshot of the index as accessed Nov. 10, not a claim that Fable is best for all tasks. The benchmark’s cost-per-task values derive from its own test and should not be confused with Anthropic’s per-token list prices.
Anthropic also recommends that users start with Claude Opus 5.5 for most coding and knowledge work and move to Fable when a task calls for it, according to the source material. In Artificial Analysis’s comparison, Opus 5.5 at Max scores 59 at $10.04 per index task, while Fable 5.2 at High scores 62 at $9.08. Those results offer a benchmark comparison, not a guarantee that Fable will be cheaper for every user’s workload.
What the Benchmark Cannot Show
The supplied results do not establish how closely the Intelligence Index’s average costs match a particular company’s real usage. Actual API bills depend on the input and output tokens a workload generates, the effort setting selected and any use of fast mode. The source does not give a task-by-task breakdown showing which evaluations drive the overall cost estimates.
The benchmark also does not show whether the score gains produce enough practical benefit to justify the extra expense for a specific user. Scores are rounded as displayed by Artificial Analysis, and the figures are a dated snapshot accessed Nov. 10. Anthropic’s claims about safety, speed and prerelease achievements are company statements; they are not established by the cost table. The source material does not provide an independent audit of those claims or later benchmark updates.
Watch for Updated Benchmarks
The immediate next step for buyers is to compare their own tasks against the published benchmark rather than treating its average as a quote. Teams evaluating Fable can track token use, elapsed time and output quality at different effort settings on a representative sample of work, then compare those results with their current model and budget. The source material does not identify a scheduled follow-up benchmark or a date for another model release.
Artificial Analysis’s rankings and costs may change as its index or tested models change. Any later comparison should be checked against its test scope, access date and pricing assumptions. Anthropic’s published per-token rates and any updates to its product guidance are also relevant to a practical cost estimate.
Key Questions
What is the cheapest reported Fable 5.2 effort level?
Low effort has the lowest reported average cost, at $3.10 per task on Artificial Analysis’s Intelligence Index. Its score is 54, and its average task time is 6.5 minutes in the reported test.
How much does Fable 5.2 cost at Max effort?
Artificial Analysis reports an average of $26.54 per index task at Max, with a score of 65 and an average task time of 46.4 minutes. That is a benchmark estimate, not a fixed price for every user request.
Which effort level offers the strongest reported balance of score and cost?
The source does not identify a universally best setting. In the reported index, High scores 62 for $9.08 per task, while Max scores 65 for $26.54. Which is preferable depends on the task, required quality, latency and budget.
Do all effort levels have different token prices?
No. The source reports the same standard rates at each effort setting: $8 per million input tokens and $40 per million output tokens. The benchmark’s estimated task cost varies because the settings use different amounts of tokens.
Are the benchmark results a guarantee of real-world performance?
No. They describe Fable 5.2 on Artificial Analysis’s nine-evaluation Intelligence Index. Scores, costs and task times on an individual company’s workload may differ, and the figures do not establish how much practical value a higher setting will provide for a particular job.
Source: ThorstenMeyerAI.com
