Ramp Expands AI Token Spend Tracking to Manage Surging AI Costs

Ramp has expanded its AI Token Spend Management product to let finance teams manage spending across AI services in an integrated way. Previously, finance teams only tracked labor costs and external vendor costs, but AI token spend is a new third category that is difficult to track.

The problem is that token costs are usage-based, spread across multiple teams and providers, tied to API keys, and invoices don't specify actual usage. Ramp's new solution collects costs from OpenAI, Anthropic, Google Gemini, Cursor, OpenRouter, and more, consolidating them into a single dashboard.

The features offered include filtering by provider, team, individual, project, and API key; setting spend limits by team, project, and key; and alerts before usage limits are exceeded. Weekly briefings present spending trends and opportunities to switch to cheaper models. A third of customers are in situations where they could use lower-cost models while handling the same tasks.

AngelList Advisors discovered a prompt caching optimization technique through the weekly briefing and resolved a $10,000 monthly loss the same day. Ramp developed this product in an environment where 99.5% of its own employees use AI tools daily, and it is offering the product for free to both existing customers and new users.

Founded in 2019, Ramp raised $750 million in a Series F funding round last June led by ICONIQ Capital, GIC, and Ontario Teachers' Pension Plan, reaching a valuation of $44 billion. Total cumulative funding has exceeded $3 billion.