• benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.

If you're building agents that read dense charts or documents, giving the model a zoom/crop tool can more than double accuracy — the Chartography numbers show tool-augmented perception beats raw model vision, with a cookbook to replicate it.

• postClaude's ant CLI gains `ant apply` for declarative agent infrastructure
• postAnthropic open-sources its commerce agent blueprints
• postAnthropic previews Function Hooks as a deeper Claude Code plugin extension point
• postFable 5.1 is now live in Claude Code and the Claude Platform.

• postFable 5.1 leads the index and still costs 20% more per taskArtificial Analysis
• postAnthropic ships Fable 5.1 with a 75% cache read price cutClaude
• postTeaching multimodal models to zoom internally, without the tool callAnt Open Source

Checking sign-in…

Loading comments…