OPEN INFRASTRUCTURE · FOR LARGE-SCALE AI
A research commons
for large-scale AI.
Shared compute, routed models, and open tools — so anyone serious about AI research can use them and contribute back.
Free for academic and independent researchers.
WHAT IT IS
Three pieces, designed to be used together.
Routed model access
One key, one client, every major model provider — OpenAI- and Anthropic-compatible, multi-tenant, with per-org budgets.
Learn more →Workspaces for research groups
Organizations, invites, and keys built for labs. Manage upstreams and budgets together; researchers move between projects without re-keying.
Learn more →Open code, open governance
The infrastructure is open source. Researchers can run upstreams, propose policy, and extend it. No private layer.
Learn more →
HOW IT WORKS
Manage in the dashboard. Call from anywhere.
| Name | Provider | Status |
|---|---|---|
| openai-default | OpenAI | Active |
| anthropic-prod | Anthropic | Active |
| fireworks-r1 | Fireworks | Idle |
import os
from openai import OpenAI
client = OpenAI(
base_url="https://api.research-computer.org/v1",
api_key=os.environ["RC_KEY"],
)
client.chat.completions.create(
model="claude-sonnet-4-6",
messages=[{"role": "user", "content": "Hello"}],
)OpenAI- and Anthropic-compatible endpoints. Drop in, keep your tooling.
PRINCIPLES
Built like scientific infrastructure, not a product.
Openness over opacity.
All code, configs, and policy are public. Governance happens in the open, not behind a contract.
Pooled, not owned.
Resources are shared by the community. One team's project doesn't block another's; budgets are visible to everyone in a workspace.
Built to be replaced.
When better infrastructure exists, ours steps aside. The goal is research, not the project itself.
Join the commons.
Free for academic and independent researchers. Talk to us about lab-scale deployments.