Claude Fable 5.1 Is Now Available for Assessments
September 1, 2026 · InfraScout Team
Claude Fable 5.1 is now selectable on Anthropic connections, at the top of the model catalog. One million tokens of context, 128k of output, and adaptive thinking across all five effort levels.
Where it sits on cost
Fable 5.1 holds Fable 5's input and output rates, and cuts cached input to a quarter of what Fable 5 charges — $0.25 per million tokens rather than $1.00.
That is the number that matters most for assessment work. A long investigation re-reads its own accumulated tool output on every turn, so cached input is the majority of what a serious conversation spends. Between the two Fable tiers, 5.1 is the one to pick.
It remains a premium tier. For quick questions, inventory lookups and routine checks, the faster and cheaper models are still the right answer — see Choosing a model for how the catalog lays out.
Opt in per connection
Adding a model to the catalog never moves an existing connection onto it. Nothing changes for your connections unless an administrator selects it.
No default moved either: the model a new connection starts on, and the models InfraScout uses internally for planning and short answers, are all where they were. Putting any of those on a premium tier is a spend decision rather than a step in supporting a new model.
Everything you rely on works with it
Adaptive thinking, the live task budget, long-run context compaction, Task Agents and agentic workflows all behave as they do on the rest of the catalog. We verify each model's capabilities against the live API when we add it, rather than inferring them from release notes, and Fable 5.1 was no exception.
To use it, an administrator picks Claude Fable 5.1 on an Anthropic connection under AI Provider Connections, then choose that connection at the start of your next deep investigation.
Questions or feedback? Reach us at info@infrascout.cloud.