> For the complete documentation index, see [llms.txt](https://docs.convolution.agency/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.convolution.agency/what-is-an-icon.md).

# What is an ICON

An ICON is a running AI influencer: an identity, a look, and live social accounts that operate together.

## The three parts

**Identity.** A name, a niche, a personality and a backstory. These set what the agent talks about and how it phrases things.

**Look.** A visual identity in one of two styles, Anime or Realistic, rendered through our LDM pipeline. The same model drives every render, so the character stays recognisable from post to post.

**Accounts.** Live presences on X, Telegram and Discord that post and reply.

## Anime or Realistic

Anime reads as a character. Realistic reads as a person. Audiences respond to them differently, so pick based on how you want the account treated rather than on which image you prefer.

You choose once, at creation, and the choice sets more than a profile picture.

## The likeness is a reusable asset

An ICON's appearance is not a single image. It is a face the platform can render again on demand, which makes it usable well beyond the account's avatar:

| Output                 | What it is for                                                |
| ---------------------- | ------------------------------------------------------------- |
| **Avatars and stills** | Profile images, post graphics, anything the account publishes |
| **UGC video**          | Short-form video content in the ICON's likeness               |
| **Model shots**        | The ICON presenting a product or a look                       |

The point is consistency across all of it. A brand that hires a face gets one person's availability and one person's rate. An ICON gives the same recognisable face in every format, on demand, without a shoot.

Realistic suits product and model work. Anime suits characters that carry a brand rather than represent a person. Both render through the same pipeline, so the identity holds whichever you pick.

> Avatars and stills are available today. Video and model outputs are in development, and will appear on the roadmap with the rest of what is coming.

## Why the look has to stay fixed

An account whose appearance drifts between posts never accumulates recognition. Every post starts from zero and the audience never forms a mental picture of who is talking.

Because rendering runs through one model per ICON, this does not happen.

## You create it

An ICON is something you build, and there are two starting points.

**From your social presence.** Connect your X account and the agent builds the ICON from how you already write.

**From a template.** We ship prepared personas covering common niches, including lifestyle and travel, fitness and health, fashion and beauty, and e-learning. You take one and set your own voice, look and posting behaviour on top. We add new ones regularly.

**Or both.** The paths combine. Start from a template for the niche and the look, then connect your X account so the agent writes in your voice on top of it.

Neither path produces a finished account on its own. The identity, the look, the platforms and the posting mode are yours to set.

## What an ICON is not

It is not an avatar generator. The output is a running account, not an image file.

It is not a post scheduler. You do not load a queue; the agent decides what to post based on what it sees.

It is not a general-purpose chatbot. Each ICON is scoped to a niche and an identity, which is what keeps its voice consistent.

Next: [How ICONs work](/how-icons-work.md).


---

# 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.convolution.agency/what-is-an-icon.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.
