About
About NeverMetered
Flat-rate, OpenAI- and Anthropic-compatible inference for coding agents. No token meter and no 5-hour or weekly limits; your plan sets how many agents run at once.
Who runs it
NeverMetered is run by Keep Pushing Forward Ltd, Offices in the UK.
How it works
- Our own GPU servers
- Requests run on dedicated GPU servers we rent and operate. At the busiest moments some requests can be answered by an overflow provider, as the Privacy Policy explains.
- One open model
- Every plan uses qwen3.8-27b, an open-weights model, through OpenAI- and Anthropic-compatible APIs.
- Flat pricing
- Plans set how many requests generate at the same time, never how many tokens you use. A GPU costs the same whether it's busy or idle, so that's what we charge for.
- Your prompts stay yours
- We never use your content to train models, and we keep only request metadata such as token counts, timings and status.
Get in touch
Questions about the service or your account: legal@nevermetered.com. The exact model build, engine, speed and data handling are on How it runs, and live service health is on the status page.
The Terms, Privacy Policy and Acceptable Use Policy set out exactly what we do with your data and account.