Ramp Enters the AI Inference Arena with Router
Corporate expense management platform Ramp is following in the footsteps of Stripe by establishing a toll booth for AI inference. On Wednesday evening, Ramp launched its own AI model routing service, dubbed Router, which allows users and companies to access and switch between various large language models (LLMs) through a unified API. According to the company, Router has been internally powering Ramp's own AI needs for the past three years.
Pricing and Availability
Router is currently available only in the United States. It is free to use for the remainder of 2026, with users only responsible for AI model inference costs. Ramp is also offering a $26 credit as a launch incentive. The company has not yet disclosed pricing for 2027.
How Router Works
Functionally, Router operates similarly to OpenRouter, though the latter offers a broader selection of AI models. Router provides access to models from OpenAI, Anthropic, DeepSeek, Moonshot, Minimax, Nvidia, xAI, and Z.ai. It also includes several “strategies” that let customers route AI requests based on their preferences. For example, users can prioritize model providers’ flex usage tiers or have Router select models based on up to three user-defined benchmarks. Additionally, users can opt to send only complex problems to premium models or test different models without manually switching.
Dashboard and Data Retention
Router includes a dashboard that displays token spend, cost, latency, fallback attempts, and other key metrics. Notably, Router has an opt-out data retention policy: by default, it records model inputs, outputs, and tool calls for one year. However, Ramp states that it removes personally identifiable information before using that data to improve the product.
Strategic Implications for Ramp
Entering the model routing business presents Ramp with a dual opportunity. First, it taps into the rapidly growing AI inference market. Second, it offers existing clients a model routing service that integrates seamlessly with Ramp’s current products, including AI token usage monitoring and spend management.
If Router becomes as popular as OpenRouter as a model testing ground, Ramp could build lasting relationships with AI labs and inference providers worldwide. This would help Ramp—which raised $750 million at a $44 billion valuation in June 2026—attract new customers and create an additional entry point for selling its expense management products.
via TechCrunch AI