Anthropic has launched a suite of reporting tools aimed at giving enterprise customers a more granular view of what their token spending buys. The announcement follows financial disclosures showing the company recorded an operating profit in Q2 and 900% net revenue retention — spending from the same customers in Q3 2025 grew nine times in Q3 2026. The new reports combine internal analytics with employee surveys, letting managers see a team’s work over a given date range, what the team produced, how much it cost, and where workflows broke down. Anthropic says it will not use this data for model training by default, and the reports also include recommendations for improving results.

The move is a clear attempt to demonstrate return on investment for Anthropic‘s tools, a major concern among large customers. Anthropic is also introducing pooled group budgets so teams can better manage token spending. The challenge is real: Jellyfish, a software developer tooling company, released a report based on 276,000 customers using agentic coding tools. Claude Code and other Anthropic-based tools remained the most widely used, accounting for nearly a third of all proposed code changes submitted for review. Spending on those tools has grown 10x since the beginning of the year, now equal to about 2% of total engineering cost. Jellyfish‘s chief scientist told TechCrunch that the 10x growth is “growth that most CFOs didn’t plan for,” and that companies now have to optimize AI spending rather than just spend.

The reporting tools also speak to switching costs. OpenAI’s Astra model hasn’t slowed Anthropic‘s adoption, according to Jellyfish data, despite perceptions on X that Codex is better. Jellyfish‘s AI product lead said Anthropic has strong enterprise motions and an incumbent advantage: workflows are standardized around it, and there’s not enough reason to switch. He also noted the investment all major AI labs are making in forward-deployed engineering teams, which embed AI consultants inside customer organizations; Anthropic‘s reporting tools can be seen as a lighter version of that hands-on help. Once an enterprise has invested in a lab’s workflows, switching becomes harder, since models differ slightly and the cost to switch is not zero.