Corporate expense management platform Ramp launched an AI model routing service called Router on Wednesday, August 20, 2026, allowing users and companies to access and switch between multiple large language models through a single API.
Router offers access to models from OpenAI, Anthropic, DeepSeek, Moonshot, Minimax, Nvidia, xAI, and Z.ai. The service includes several routing strategies, such as directing queries to models based on up to three user-specified benchmarks, routing only difficult problems to more expensive models, or setting preferences for providers’ flex usage tiers. A dashboard lets users monitor token spend, cost, latency, and fallback attempts.
Ramp says it has been using the technology internally for its own AI needs over the past three years. The service is available only in the United States. It is free to use for the remainder of 2026 — though users still pay underlying AI model inference costs — and comes with a $26 credit at launch. Pricing for 2027 has not been announced.
Router operates with an opt-out data retention policy, recording model inputs, outputs, and tool calls for one year by default. Ramp says it removes personally identifiable information before using that data to improve the product.
For Ramp, the move could serve two purposes: tapping the AI inference market and offering existing clients a routing service that complements its current products, which already include AI token usage monitoring and token spend management. Building relationships with AI labs and inference providers through Router may also give Ramp a new entry point for its expense management business. Ramp raised $750 million at a $44 billion valuation in June 2026.
Source: TechCrunch