Ramp has launched its own AI model routing service, Router, on Wednesday evening.
The corporate expense management firm joins Stripe in building toll houses for AI inference. Ramp states it has been using this internal router for its own AI needs over the past three years.
Access is currently limited to the United States. The service is free for the rest of 2026, though users must still pay for the underlying AI model inference costs. A $26 credit accompanies the launch. Ramp did not disclose pricing for the following year.
How it works
Router functions like OpenRouter, though the latter currently offers a wider selection of models. It provides access to systems from OpenAI, Anthropic, DeepSeek, Moonshot, Minimax, Nvidia, xAI, and Z.ai.
The platform includes strategies to route requests based on user preference. One option allows users to set a preference for model providers’ flex usage tiers. Another lets Router select a model based on up to three user-specified benchmarks. Users can also choose to route only difficult problems to expensive models or test models without switching interfaces.
A dashboard displays token spend, cost, latency, fallback attempts, and other details.
Router includes an opt-out data retention policy. It records model inputs, outputs, and tool calls for one year by default. The company says it will remove personally identifiable information before using that content to improve the product.
For Ramp, entering the model routing market offers two opportunities. It can tap the booming AI inference market and offer a service that fits with existing products, including AI token usage monitoring and spend management.
If Router proves as attractive a model testing arena as OpenRouter, Ramp may build long-standing relationships with AI labs and inference providers worldwide. That could help the company, which raised $750 million at a $44 billion valuation in June, gain new customers. It also creates a new point of entry for selling its expense management products.
What it means
Companies using Ramp now have a built-in way to compare AI providers without leaving their expense platform. The free access for the rest of the year lowers the barrier for teams to test different models and pricing structures before committing to a specific vendor.
