> For the complete documentation index, see [llms.txt](https://docs.gpu-fi.uk/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.gpu-fi.uk/gpu-infrastructure/inference-api.md).

# Inference API

Describes the planned inference API and its intended use.

{% hint style="info" %}
**Phase 4: Inference API is planned, not currently live.**
{% endhint %}

### What it is

GPUFi intends to introduce an inference API that allows developers to pay for compute. The stated model is payment per token of compute.

Technical API details have not been specified. This includes endpoints, model names, SDKs, frameworks, token prices, request limits, latency targets, availability guarantees, and providers.

### Intended economic flow

```
GPU infrastructure
  ↓
Inference API
  ↓
Developer compute usage
  ↓
Compute revenue
  ↓
Mining pool
```

This is a future Phase 4 mechanism. It is intended to create another ecosystem revenue source beyond trading tax and future GPU rental.

### Relationship to GPU rental

The intended infrastructure path has three stages:

| Phase   | Intended role                                                             |
| ------- | ------------------------------------------------------------------------- |
| Phase 2 | Treasury acquires initial GPUs                                            |
| Phase 3 | External operators contribute additional GPUs                             |
| Phase 4 | Infrastructure can also serve developer compute through the inference API |

The roadmap therefore describes two intended infrastructure revenue paths:

* GPU rental
* Developer compute

Neither revenue path exists in Phase 1.

### Relationship to mining

Future GPU rental revenue and inference API revenue are intended to contribute to the mining pool. This expands the intended funding model beyond its initial trading-tax mechanism.

```
Phase 1
Trading tax
  ↓
Mining pool

Later
Trading tax
  +
GPU rental
  +
Inference API
  ↓
Mining pool
```

Trading tax remains part of this progression. The roadmap does not state that infrastructure revenue replaces it.

### Toward greater revenue diversity

The intended model is for actual compute usage to create additional ecosystem revenue. Phase 4 is intended to move the ecosystem toward greater revenue diversity and reduced dependence on trading activity alone.

This outcome depends on future infrastructure deployment and use. Continue to the [Roadmap](/roadmap.md) for planned milestones.


---

# 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 by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.gpu-fi.uk/gpu-infrastructure/inference-api.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

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.