Introduction
Understand Ando, inference, and how the platform works.
Ando brings open models into the apps and agents people return to every day. It keeps inference clear, grounded, and close to practice, with account-owned access and a visible usage trail.
Inference is the moment an app asks a model to help shape a response. In Ando, that moment stays connected to an account, a connection, and a clear boundary before the response returns.
Ando works as a calm layer between the app in front of you and the models behind it. Apps connect through Virtual Keys or supported agent payment flows. Ando keeps access, model choice, and usage legible before returning the model response.
Ando keeps inference close to everyday practice, while giving each request a clear account, connection, and usage trail.
How Ando fits together
Your account keeps inference access, app connections, and model use visible in one place.
A connection links one app, agent, tool, or flow to the boundary it should use.
Most apps use a scoped Virtual Key. Accountless agent paths can use supported payment flows.
Requests move through Ando for model access, clear routing, usage reporting, and boundaries.
What this section covers
Move from account setup to your first working connection.
02AppsChoose from supported app presets and follow app-specific setup steps.
03SettingsKeep model use deliberate, visible, and inside the boundary you choose.
04Inference TokensUnderstand how model work becomes visible usage and reported spend.
Core objects
The product, agent, interface, or everyday flow that sends model requests through Ando.
The governed relationship between your Ando account and one app or flow.
The authorization token format Ando uses for scoped, boundary-aware access.
The credential most connected apps use when they ask models through Ando.
These objects work together: choose an app preset, create a connection, reveal the Virtual Key, then configure the app to send requests through Ando.