> For the complete documentation index, see [llms.txt](https://docs.vapinetwork.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.vapinetwork.ai/router/how-router-works.md).

# How Router works

Router is coming soon; its planned OpenAI-compatible endpoint will route and meter requests to upstream models.

{% hint style="info" %}
vAPI Router is coming soon. This page describes how it works at launch.
{% endhint %}

vAPI Router is a hosted, OpenAI-compatible inference endpoint at `router.vapinetwork.ai`. Use it when you want one interface for models served by vAPI's upstream providers.

## One OpenAI-compatible interface

Use the Router base URL with an OpenAI-compatible SDK or HTTP client. The SDK base URL includes `/v1`, as shown in [Call your first model](/router/api-quickstart.md).

Router hosts five model kinds:

| Kind          | Deployments in the catalog | Metering unit    |
| ------------- | -------------------------: | ---------------- |
| Text          |                        101 | Per token        |
| Embedding     |                          9 | Per token        |
| Speech        |                         11 | Per character    |
| Transcription |                          5 | Per audio second |
| Image         |                         24 | Per image        |

The current source catalog contains 150 deployments. Counts can change when the upstream catalog or catalog policy changes. See [Models](/router/models.md) for the catalog policy.

## Compute and Router balance

Compute is the daily usage entitlement a wallet earns from staking vAPI. Each Router key spends one source selected at creation: Compute or balance. CLI chat, local MCP chat, and SDK `client.router.chat()` try Compute first, then a stored balance key after Compute is exhausted. Direct OpenAI-compatible clients spend the source attached to the supplied key and must switch keys themselves.

Linked agents receive a separate daily Router allowance when the owner approves the link. An agent's Compute use is bounded by both that allowance and the owner's daily Compute limit. See [Use Router from your agents](/router/use-router-from-your-agents.md) and [Buy Router balance](/router/buy-router-balance.md).

## Upstream

Venice is the upstream provider behind Router. Router connects its OpenAI-compatible interface to that upstream.

## What vAPI adds

vAPI provides the control plane around the proxy:

* wallet-bound Router keys
* daily Compute usage limits
* purchased Router balance
* usage and model views
* catalog policy
* the console and public API

The data plane is the open-source LiteLLM Proxy. vAPI writes the control plane and does not write proxy code. The models are served by the upstream provider.

For key creation and wallet access, see [Access and keys](/router/access-and-keys.md). For a working request, see [Call your first model](/router/api-quickstart.md).

Checked on 2026-10-02.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:

```
GET https://docs.vapinetwork.ai/router/how-router-works.md?ask=<question>&goal=<endgoal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.