Hosting providers
Run AI assistants as a service
Offer a self-hosted AI assistant as a paid product to many customers, each isolated and independently billed.
The problem
- You want to resell managed AI assistants, but a single-user gateway means one deployment per customer — an operational nightmare at any real scale.
- You need to bill each customer for their actual LLM usage, and prove it, without hand-rolling a metering pipeline.
- Customers churn, migrate, and demand data portability — you need clean snapshot and restore, not a database surgery every time.
With OpenClawMU
- One gateway serves many tenants. Per-tenant cost accounting records token usage against each tenant, and the report generator rolls it into per-period totals against your rate card.
- S3-compatible backup/restore produces tenant-portable snapshots (AWS S3, MinIO, GCS, R2) that are path-traversal hardened — a tenant migrates to another gateway with one backup and one restore.
- Quotas enforce plan limits at the request boundary; an over-quota tenant gets a 429 with the next-window timestamp instead of eating your margin.
How it works
Getting there, step by step.
- 1
Define your rate card
Configure the billing rate card in the gateway config. Token usage is priced per tenant against it.
- 2
Provision and price tenants
Create a tenant per customer and attach a quota that matches their plan.
- 3
Generate usage reports
Run `openclaw billing report <tenant> --period current-month --csv` to produce an auditable per-tenant invoice input.
- 4
Offer portability
Back up a tenant to your S3 bucket on demand; restore it on another gateway to migrate or to satisfy a data-export request.
# one gateway, many tenants
npm install -g openclaw@latest
openclaw onboard --install-daemon
# a tenant per customer / team
openclaw tenants create acmeFrequently asked
FAQ — run ai assistants as a service
Every question here is also emitted as FAQPage JSON-LD — so Google AI Overviews, Perplexity, and ChatGPT browsing extract the answers cleanly. Crawler-friendly fallback below for screen readers and bots that don't run JS.
Every model call records a token-usage row scoped to the tenant. Cost snapshots multiply token counts by your configured rate card into per-period totals, exportable as CSV.
Any S3-compatible store — AWS S3, MinIO, Google Cloud Storage, or Cloudflare R2. Backup and restore targets are path-traversal hardened.
Related use cases
-
Ship a multi-tenant bot platform
The agent platform you're building needs isolation between customers without inventing it from scratch.
Read → -
Give internal teams a shared gateway
One internal gateway your departments share — isolated tenants per team, admin-controlled, no data crossover.
Read → -
How isolation works
The load-bearing change vs. upstream — tokens, per-tenant directories, path-traversal hardening.
Read the deep dive →
EXFOLIATE!
Run your own gateway today.
Apache-2.0, self-hosted, no SaaS layer between you and your users. Install the CLI, create your first tenant, mint a token — you're routing traffic in 60 seconds.