CompanyFabric vs LiteLLM
LiteLLM gives you a proxy server and SDK you run yourself — your infra, your provider keys, your ops. CompanyFabric is the hosted version of the same idea, with metering, prepaid billing, and a model library maintained for you.
The differences
Side by side
| CompanyFabric | LiteLLM | |
|---|---|---|
| Hosting | Fully hosted — no proxy to deploy or scale | Self-hosted proxy you deploy, upgrade, and monitor |
| Billing & metering | Built in: exact per-call cost, prepaid balance, spend caps per key | Bring your own — budgets exist, invoicing doesn't |
| Provider keys | Ours (credits) or yours (BYOK at 0%) | Always yours |
| Media models | Video/image/audio metered natively | Text-centric; media support varies by provider |
| Cost | Pay per call; no infrastructure | Free software + your compute + your time |
Verdict
Which should you pick?
Run LiteLLM when policy requires every byte on your infra and you have the ops capacity. Pick CompanyFabric when you'd rather change a base URL than operate a proxy fleet — or use both, LiteLLM pointing at us as an upstream.
FAQ
Frequently asked questions
Can LiteLLM route through CompanyFabric?
Yes — add us as an OpenAI-compatible provider in your LiteLLM config with base_url api.companyfabric.com/v1. Teams do this to keep LiteLLM's routing rules while consolidating billing.
One key. Every model. Exact prices.
Migrating from LiteLLM is a base-URL change. Free starter credits included — no subscription required.