# ECU credits — concept

ECU stands for **Evaluation Compute Unit**. It's the unit Stratix uses to meter compute-intensive operations. Premium tier is pay-as-you-go on ECU; Free tier ships with a starter balance.

## What ECU bills

* **Model inference during evaluations** — prorated by token, by model
* **Judge runs** — judges are LLMs; they consume ECU per-call
* **GEPA optimization** — per-iteration
* **Trace evaluation runs** — trace × judge calls

## What ECU does NOT bill

* Browsing the catalog (Public or Premium)
* Configuring evaluations, spaces, judges, scorers
* Saving and sharing
* API key generation
* Tokens consumed by your **own provider key** (BYOK custom models bill against your provider, not ECU)

## Why a unit-based model

A unit-based model lets Stratix offer:

* Pay-as-you-go without complicated multi-provider billing
* Cost previews before runs (token estimates × per-model ECU rates)
* Per-team and per-project consumption tracking

## Cost preview

Before any compute-intensive job, Stratix shows worst-case ECU consumption. You approve before it runs. Real consumption may be lower if a job short-circuits.

## ECU vs your provider tokens

* **Public model** evaluations bill against ECU (the platform pays the provider)
* **BYOK custom model** evaluations bill against **your** provider (the model tokens) but ECU is still consumed for the platform engine work

## Volume pricing (Enterprise)

Enterprise contracts include negotiated volume ECU pricing. Talk to sales when monthly consumption justifies it.

## Where to next

* [ECU credits and billing (Premium)](/11.-admin/ecu-credits-billing.md)
* [BYOK models](/5.-select-pick-the-model/byok-models.md)
* [Pricing](/11.-admin/11-admin.md)


---

# Agent Instructions: 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:

```
GET https://docs.layerlens.ai/11.-admin/concepts-ecu-credits.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
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.