Claude Fable 5.2 brings sharper judgment to messy enterprise data
Blog post from Box
Anthropic reports that Claude Fable 5.2 outperforms Fable 5.1 on Box AI Complex Work Eval tasks involving enterprise spreadsheets, PDFs, presentations, and images, particularly in consumer products, energy, public sector, and technology scenarios. Evaluated against dozens of weighted criteria across repeated runs, the model improved scores by identifying source-data quirks before performing analysis, incorporating relevant information from across full file sets, and presenting conclusions first with explicit assumptions and judgment calls. Examples include correcting spreadsheet timestamps stored slightly before the hour, inferring grade categories encoded by cell color, applying a later case-manager correction that changed a school-program recommendation, and selecting an earlier product-launch option consistent with scenario constraints. Anthropic says the model also provides traceable outputs by linking results to source control totals, policies, and thresholds, while operating about 10% faster than Fable 5.1, and plans to make it available soon through Box AI.
No tracked trend matches for this post yet.
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.