Ramp has launched Router, an AI model routing service that lets companies access and switch between multiple large language models through a single API. The expense management company says it has used the underlying routing system internally for about three years to manage its own AI workloads.
Router is currently available only in the U.S. and will be free to use through the end of 2026, although customers will still pay the underlying AI inference costs. Ramp is also offering a $26 launch credit and has not yet disclosed pricing for 2027.
Router Supports Models From Several AI Providers
The service offers access to models from OpenAI, Anthropic, DeepSeek, Moonshot, Minimax, Nvidia, xAI, and Z.ai. It works similarly to OpenRouter, although OpenRouter currently supports a larger selection of models.
Customers can set routing rules based on their priorities. One strategy can favor lower-cost flex usage tiers, while another can choose models according to as many as three user-selected benchmarks.
Router can also send only more difficult requests to higher-cost models or let developers test different models without manually changing their integrations. Its dashboard shows metrics including token usage, cost, latency, and fallback attempts.
Ramp Stores Model Activity for One Year by Default
Router uses an opt-out data retention policy. By default, Ramp says it will store model inputs, outputs, and tool calls for one year.
The company says it will remove personally identifiable information before using stored content to improve the product. Customers who do not want that retention will need to opt out.
The service fits alongside Ramp’s existing tools for monitoring AI token usage and managing related spending. Model routing gives the company another way to connect AI infrastructure costs with the financial management products it already sells to businesses.
Ramp Expands Further Into AI Infrastructure
Ramp’s entry into model routing also gives it exposure to the growing AI inference market. If developers use Router to compare and test models, the service could also give Ramp closer relationships with AI labs and inference providers.
The launch comes after Ramp raised $750 million in June at a $44 billion valuation. It also follows Stripe’s recent expansion into AI routing infrastructure.
Ramp could use Router both as a standalone AI infrastructure product and as another entry point for customers that may later adopt its broader expense management services.
Featured image credits: Ramp
For more stories like it, click the +Follow button at the top of this page to follow us.