Unlock inference for OpenClaw
Intelligent SLA-aware routing that matches each workload to the right compute topology.
"We make it possible for the 90% of agentic workloads that are routine and always on, to run on leaner and often underutilized compute topologies"
— Behnam Bastani, CEO
The infrastructure assumption has changed.
When your inference layer is rented, one policy update is all it takes to break your operations. Anthropic's restrictions on using OpenClaw as a tool within Claude exposed how fragile single-provider dependency really is.
OpenInfer was built before this moment. But this moment proves why it matters.
Enterprises running OpenClaw on OpenInfer continue operating without modification — workloads route to what is available, what is compliant, and what is cost-effective, automatically.
Get Started
Accept the license, then configure OpenClaw with one of two options below.
Free version routes to low SLA.
Beta License Agreement
OpenInfer, Inc.1. Acceptance. By clicking "I Agree", you agree to be bound by this Agreement.
2. Scope. Access to the inference stack solely for internal evaluation during Beta. Third-party software licensed separately.
3. Beta Program. Experimental, may contain errors, may be discontinued. Not for production use.
4. No Support. "As-is" and "as-available" with no guarantees.
5. Acceptable Use. Lawful purposes only.
6. Restrictions. No copying, reverse engineering, reselling, or publishing benchmarks.
7. Data. You are responsible for all data processed.
8. IP. OpenInfer retains all rights.
9. Disclaimer. NO WARRANTIES OF ANY KIND.
10. Liability. Total liability shall not exceed $0.
11. Termination. Access may be terminated at any time.
12. Governing Law. California. San Mateo County courts.
Accepted
Option 1: Replace your OpenClaw config
Download or copy the full config file to replace your existing configuration.
{
"models": {
"providers": {
"local": {
"baseUrl": "https://api.openinfer.io/openai/v1",
"apiKey": "9b8f3d7c1a6e4f20b5c9a1d8e7f34c62",
"api": "openai-responses",
"models": [
{
"id": "@oi/beta",
"name": "@oi/beta",
"reasoning": false,
"input": [
"text"
]
}
],
"headers": {
"X-Session-ID": "openclaw-responses-2"
}
}
}
},
"agents": {
"defaults": {
"model": {
"primary": "local/@oi/beta"
},
"workspace": "~/.openclaw/workspace",
"timeoutSeconds": 60000
}
},
"tools": {
"profile": "coding",
"web": {
"search": {
"enabled": true,
"provider": "duckduckgo"
}
}
},
"commands": {
"native": "auto",
"nativeSkills": "auto",
"restart": true,
"ownerDisplay": "raw"
},
"gateway": {
"mode": "local",
"auth": {
"mode": "none"
}
}
} Option 2: Manually update your existing config
Add a provider to your existing ~/.openclaw/openclaw.json:
1. Add a new provider
{
"baseUrl": "https://api.openinfer.io/openai/v1",
"apiKey": "9b8f3d7c1a6e4f20b5c9a1d8e7f34c62",
"api": "openai-responses",
"models": [
{
"id": "@oi/beta",
"name": "@oi/beta",
"reasoning": false,
"input": [
"text"
]
}
],
"headers": {
"X-Session-ID": "openclaw-responses-2"
}
} 2. Set the default model
openinfer/@oi/beta
Sign up for adaptive SLA routing
Free version routes to low SLA. Sign up to get access to adaptive SLA routing.
You're on the list — we'll be in touch.