CLAUDE LABJP
COST — v2.1.239 folds the 1.1x US-only-inference premium for data-residency workspaces into /cost, the status line, and max-budget-usd estimatesBUDGET — Until now those workspaces saw estimates that ran below the real invoice. If you set a spend cap, it is worth reconciling this month's bill against your /cost historyFIXES — v2.1.240 is a fixes-and-reliability release, continuing recent work on MCP v2 connections, notification hooks, idle sessions on Linux, and Windows path validationACADEMY — Claude Academy is now open: a learning hub for safe, effective AI use, with courses, tutorials, badges, and personalized recommendationsPLATFORM — The computer use tool, browser use tool, Skills API, and Files API are now available on the Claude Platform, opening a path to move skill workflows off your local machinePRICING — Claude Sonnet 5's introductory $2 per million input and $10 output ends August 31, with standard $3 and $15 pricing from September 1. Eight days to goCOST — v2.1.239 folds the 1.1x US-only-inference premium for data-residency workspaces into /cost, the status line, and max-budget-usd estimatesBUDGET — Until now those workspaces saw estimates that ran below the real invoice. If you set a spend cap, it is worth reconciling this month's bill against your /cost historyFIXES — v2.1.240 is a fixes-and-reliability release, continuing recent work on MCP v2 connections, notification hooks, idle sessions on Linux, and Windows path validationACADEMY — Claude Academy is now open: a learning hub for safe, effective AI use, with courses, tutorials, badges, and personalized recommendationsPLATFORM — The computer use tool, browser use tool, Skills API, and Files API are now available on the Claude Platform, opening a path to move skill workflows off your local machinePRICING — Claude Sonnet 5's introductory $2 per million input and $10 output ends August 31, with standard $3 and $15 pricing from September 1. Eight days to go
TAG

retrieval augmented generation

1 articles
Back to all tags
Related:
Claude API1RAG1vector search1groundedness1evaluation1TypeScript1
API & SDK/2026-07-09Advanced

When the RAG Started Being Confidently Wrong — Field Notes on Measuring Retrieval Misses With Groundedness

In a Claude API RAG, the answers stay fluent while the facts drift. Often the cause is a silent recall decay on the retrieval side, missing the document that holds the answer. Field notes on measuring groundedness and retrieval hit rate and walking the system back, with working code and real numbers.