Machina AI Router · Route
AvailableRoute requests through configured profiles.
Define approved provider profiles and route requests by task, latency, data handling, and workflow requirements.
- 01Profile. Define approved provider and model profiles for the deployment.
- 02Constrain. Attach task and operating requirements to the request path.
- 03Route. Send the request through the configured profile appropriate to the job.
- 04Observe. Inspect the resulting activity and use Arena when a structured comparison is needed.
The job
What Machina AI Router does
The Router separates a product's model-facing request path from one hard-coded model choice. Teams configure providers and profiles for their requirements, then route requests according to task and operating constraints.
- 01
Configured providers
Use the providers and profiles approved for the deployment.
- 02
Task-aware routing
Route requests according to the job rather than one universal model default.
- 03
Operating constraints
Keep latency, data handling, and workflow requirements part of routing design.
- 04
Evaluation connection
Use Arena evidence to inform configuration decisions.
What does the Machina AI Router route?
It serves model requests through configured providers and profiles according to the task and its operating constraints.
How do I integrate the AI Router?
Use the documented interfaces available for your deployment. This page does not introduce a public OpenAI-compatible endpoint.
Machina AI Router in the stack
How AI Router connects to the wider platform.
Choose a flow, then select a component to see how the work moves.
Use the component buttons to explore the architecture.
Select a component
Run flow
Put your context to work.
Studio runs the workflow, Router serves configured model requests, and output follows the delivery policy, with review when required.
Selected component
Machina AI Router
Route model requests through providers and profiles configured for the task and its operating constraints.
Explore Machina AI Router →View connections
- Studio → Router: Model request
- Router → Models: Request to configured model
- Models → Router: Model response
- Router → Studio: Response to workflow
Routing example
Different model profiles for extraction and editorial drafting
A workflow routes structured extraction and audience-facing drafting through separately configured profiles while preserving its source and review rules.
Inputs
- Task type
- Approved provider profiles
- Latency and data-handling constraints
Output
A model response served through the configured route, inside the surrounding workflow controls.
One platform