Objective-aware selection
Choose balanced, quality, cost, latency, private, sovereign, or local-only objectives.
One endpoint.
Every eligible intelligence layer.
Keep the OpenAI client pattern you already know. Change the base URL, use the MC-1 model alias, and let policy-aware routing handle the rest.
QUICKSTARTFrom zero to
From zero to
first request.
The primary chat-completions interface is OpenAI-compatible. Production access still requires valid MC-1 credentials and an eligible project.
from openai import OpenAI
client = OpenAI(
base_url="/v1",
api_key="$MC1_API_KEY"
)
response = client.chat.completions.create(
model="colomboai/mc-1",
messages=[{"role": "user", "content": "Build this..."}]
)Pin eligible routes
Select models and providers explicitly when deterministic control matters more than optimization.
Inspect every decision
See route, policy, usage, evaluation, and operational evidence in the console.
OPEN ECOSYSTEMBuild the intelligence layer
Build the intelligence layer
with us.
Infrastructure providers and model labs can become eligible participants after integration, contract validation, and evidence-backed evaluation.
Become an inference provider.
Make approved capacity eligible for intelligently routed workloads.
Integrate infrastructure →Add a model to MC-1.
Submit a model to the Model Intelligence evaluation pipeline.
Submit a model →Integrate once.
Use one compatible interface while retaining route observability and explicit controls.
Start in the console →MODEL ALIAS
Open Console ↗colomboai/mc-1
Primary endpoint: /v1/chat/completions