Anthropic is introducing reporting and budgeting tools aimed at helping enterprise customers understand what they are getting from their AI spending.
The reports will let managers examine a team’s activity over a selected period, including what it produced, how much its work cost in tokens and where workflows encountered problems. Anthropic also says the system will recommend ways to improve results. The company says customer data from these reports will not be used to train its models by default.
A push to prove AI’s value
The release comes as companies try to move beyond broad AI adoption and determine whether their deployments are producing measurable business benefits. TechCrunch reports that Anthropic’s internal financials shared with investors showed an operating profit in the second quarter and 900% net revenue retention, meaning spending from the same customer group grew ninefold year over year in the figures cited.
Anthropic is also adding pooled group budgets so teams can manage token usage collectively. The tools are particularly relevant to coding teams, where usage of agentic systems can expand quickly and create costs that were not part of earlier planning.
TechCrunch cited a Jellyfish analysis of 276,000 customers that found Claude Code and other Anthropic-based tools accounted for nearly one-third of proposed code changes submitted for review. The report said spending on those tools had increased tenfold since the start of the year and reached about 2% of total engineering costs.
The reporting features may also help Anthropic make its enterprise relationships harder to replace. Companies often invest in standardized workflows, internal expertise and implementation support around a particular AI system, making a switch more involved than changing tools as an individual developer.