RunInfra

RunInfra, part of RightNow, hosts open models behind one API key that works with the OpenAI and Anthropic SDKs. Chat models expose OpenAI-compatible chat completions and Anthropic-compatible Messages, with tool calling, JSON mode and streaming listed as capabilities. Session-aware routing keeps repeat calls close to cached context, and cached input is billed at a lower rate. The model library lists 8 models, and a Coding plan is aimed at tools such as Claude Code, Codex and OpenCode.

Pay as you go works through prepaid credits: top-ups start at $5, spend comes from one balance, and there are no seats or per-user fees. Top-ups stay valid for one year, and new users can get $1 free to test the Model APIs. Prices are per token; for example, Ornith 1.5 122B is listed at $0.20 per 1M input tokens and $0.80 per 1M output tokens. Coding plans are Starter $10, Pro $29 and Team $99, with usage that resets every 5 hours and every week.

The Enterprise tier has custom pricing and adds audit logs and role-based access control, custom volume and contract terms, custom SLAs up to 99.99%, SOC 2 Type II attestation, and a dedicated CSM with a private Slack. On security, the site lists TLS in transit, AES-256 at rest, workspace isolation and no training on your data, and its documentation says hosted Model APIs are zero data retention by default, with prompt text never stored.

Several limits are stated on the pages. Pricing is published in USD. The comparison table says Pay as you go has no contractual SLA and does not include SOC 2 Type II, while Enterprise includes both. Cache hits are best effort and never guaranteed, even though the site reports up to 99.6% cache hit rate over the last 24 hours. At a plan limit, the choice is credits or free Standby. The terms are governed by the laws of the State of Delaware, United States.