About Boundless

Boundless Inference is the inference service for open-weight models run by Boundless Networks, Inc. You call the models through OpenAI-compatible and Anthropic-compatible endpoints, pay with prepaid credits, and can inspect usage for every request. The wider Boundless project lives at boundless.network.

The Boundless Platform Services Terms name the legal operator and govern the service. Send service questions to support@boundless.network (the support page lists what to include) and legal notices to legal@boundless.network.

What you can check#

The model catalog and rate card list the models on offer and what they cost. The API documentation covers the supported endpoints and their limits.

Privacy at Boundless explains what we record for billing and what happens to request content. The governing documents are the customer terms and the company privacy policy. Models run on third-party infrastructure, so read the limits in Section 3 of the terms before you rely on any data-handling claim for compliance.

The status badge in the site footer shows whether the service is up. Section 2.6 of the customer terms makes no service-level, uptime, or availability commitment, and the badge does not add one.

New accounts are designed to be active the moment they sign up. Support is provided "as available" under the terms.

What this site does not publish#

There is no customer case study, security certification, or leadership biography for the inference service here, and none should be assumed. For procurement questions, customer references, security documentation, or a written commitment for your use case, contact the team.

The Vercel Sandbox setup guide shows how to run coding agents in Vercel Sandbox against Boundless. It documents an integration and makes no claim that Vercel uses or endorses the service.