RunInfra

Hosted open A.I. models behind one API key that works with OpenAI and Anthropic SDKs

Visit RunInfraPaid · from $10/monthsource: runinfra.ai

✅ RunInfra is a service from RightNow that hosts open A.I. models and lets you reach them through one API key. An API is a way for one program to talk to another. This one works with the OpenAI and Anthropic SDKs, which are the toolkits many developers already use.

🔎 The pages say it's built for A.I. agents, which tend to send the same context again and again. RunInfra keeps a session warm so repeat calls go to the same replica and can be billed at a lower cached price. You can pay per token with a prepaid balance, or pick a coding plan for tools like Claude Code, Codex and OpenCode. A setup prompt lets your coding agent connect for you, and it shows you the plan and waits for your yes before it changes anything.

What it does

Works with both OpenAI-style and Anthropic-style requestssource: runinfra.ai
Repeat calls are routed to the replica holding their contextsource: runinfra.ai
Cached input is billed at a lower rate and shown in your usagesource: runinfra.ai
Models support tool calling, JSON mode and streamingsource: runinfra.ai
Coding plans reset limits every 5 hours and every weeksource: runinfra.ai
Prompts aren't stored and aren't used for trainingsource: runinfra.ai

Our take (opinion)

Best for

💻 Developers who run coding agents like Claude Code or OpenCode and want one key for open models
💸 Teams that send lots of repeated context and want lower cached-input prices
🧩 People who already use the OpenAI or Anthropic SDKs