Corporate expense management platform Ramp has introduced Router, an API service that routes requests between large language models and allows customers to switch among providers. Ramp says it has been using the same router internally for its own AI needs for the past three years.
Availability and pricing
Router is available only in the United States. The service is free to use through the end of 2026 (users remain responsible for any AI model inference charges), and new users receive a $26 launch credit. Ramp has not disclosed pricing for 2027 and beyond.
Models and routing features
The current model lineup accessible via Router includes OpenAI, Anthropic, DeepSeek, Moonshot, Minimax, Nvidia, xAI, and Z.ai. Router also offers several routing “strategies” that help customers direct requests according to preferences. For example, one strategy prefers providers’ flex usage tiers, while another chooses a model based on up to three user-specified benchmarks.
Practical uses include routing only difficult problems to higher-cost models or switching between models for testing without manual reconfiguration.
Dashboard and metrics
Users have access to a dashboard that displays token spend, cost, latency, fallback attempts, and other operational details. These metrics are intended to help teams manage costs and optimize which models handle particular workloads.
Data retention and privacy approach
Router applies an opt-out data retention policy: by default it records model inputs, outputs, and tool calls for one year. Ramp says it will remove personally identifiable information (PII) before using recorded content to improve the product.
Strategic rationale for Ramp
Entering the model routing market gives Ramp a twofold opportunity: capture demand in the expanding AI inference market, and provide an integrated routing capability to its existing customer base alongside features like token usage monitoring and spend management.
If Router becomes a popular environment for model testing — similar to how OpenRouter has functioned — Ramp could establish durable relationships with AI labs and inference providers. Those ties could drive new customer acquisition and create an additional channel to sell Ramp’s expense-management offerings.
Financial context
Ramp raised $750 million in June at a $44 billion valuation. The Router launch represents a new product direction that could extend the company’s reach into AI infrastructure and services.
