What you can build
cLLMHub is a two-sided marketplace. Hosts publish models and earn per token. Developers load credits and pay only for what they use.
For hosts
Monetize idle GPU time
Your GPU is sitting idle most of the day. Publish a model, set your per-token price, and earn whenever a developer calls it. Keep 85% of every dollar.
Sell access to fine-tuned models
You trained a specialist model — code, legal, medical, a specific language. List it on the catalog so other developers can pay per token to use it without standing up their own infra.
For developers
Pay-per-token without lock-in
Load credits, call any model in the catalog, and stop paying when you stop using it. No subscription, no minimum spend, no per-seat fees.
Compare and switch models cheaply
Try different open-weight models side by side. Routes are pluggable — change the model name in your request and you’re using a different host. No re-platforming.
Drop into existing OpenAI code
Point any OpenAI SDK at cllmhub.com/v1 and your existing app keeps working. Same shape, same streaming, same tool calls.
Power retrieval and search
Build RAG over your own documents using catalog models — embeddings, summarization, ranking — and only pay for the tokens you actually generate.
Automate background workflows
Run AI jobs from cron tasks, data pipelines, or queue workers. Per-token billing means low-volume automations cost cents, not subscription dollars.
Self-host and self-serve
You can be both host and developer on the same account. Publish a private model for your team, then call it from your app — self-consumption is always free.