Fable 5.2 leads the AA Intelligence Index, with a costly maximum setting
Anthropic’s Fable 5.2 leads Artificial Analysis’s broad capability index and cuts standard token prices by 20%, while its maximum-effort task cost remains substantially above Carbon’s.
Anthropic released Fable 5.2 and Mythos 5.2 on November 9.
They share the underlying model; Fable carries additional cybersecurity and biology safeguards, while Mythos requires organizational verification. The release is aimed at difficult, open-ended technical work. Anthropic still recommends Opus 5.5 as the starting point for most everyday work. [1][2]
Standard API prices fall from $10 to $8 per million input tokens and from $50 to $40 per million output tokens. Optional fast mode costs $16 and $80. Lower token prices do not establish a lower bill for a completed assignment: the amount of reasoning and delegated work also matters. [1][2]
Artificial Analysis gives Fable 5.2 at Max with default fallback a rounded Intelligence Index score of 65, versus 60 for Carbon at High. Its reported average cost per index task is $26.54 for Fable and $7.76 for Carbon, using Carbon’s regular API rates. These are results for particular configurations and a nine-evaluation task mix, not a universal ranking for every job. [4][5][6]
Sources & details 6 sources
Discuss in your voice app
Take the reporting, visual explanations and original sources into a conversation. Copy the brief, paste it into your preferred assistant, then start its voice mode.
Paste the brief into your voice app to discuss this reporting.Reporting sources
- Anthropic — Fable 5.2 and Mythos 5.2
November 9 launch, split safeguards, list prices and provider recommendations; launch performance claims are vendor measurements.
- Claude Platform — changes in Fable 5.2
Pricing, migration behavior, supported limits and model-specific data retention.
- AWS — Fable 5.2 model card
Hosting-provider confirmation of the November 9 launch, context/output limits and preserved-thinking behavior.
- Artificial Analysis — Fable 5.2 Max
Independent evaluator's current v5.0.1 score and weighted average task cost; configuration includes default fallback.
- Artificial Analysis — Carbon High
Same evaluator's Carbon configuration and regular-price task-cost estimate; not Google's introductory price.
- Artificial Analysis — evaluation methodology
Task mix and measurement framework; benchmark results are not a guarantee for another workload.
Images & figures
AI-generated editorial illustration of coordinated technical work; a conceptual mechanism, not a real device or model architecture.
Euangelion · Source · Usage terms
Fable 5.2 standard costs 20% less per token than 5.1; fast mode costs twice the standard rate.
Euangelion · Source · Usage terms · Download figure