# TITLES

> Creative tools powered by artist-trained AI models. Creators generate images and videos using LoRAs trained by real artists; every generation pays the artist a royalty automatically. Publish on-chain to share your creation with full provenance.

TITLES is a creative studio where creators use artist-trained AI models to generate images and videos. Every generation sends a share of the fee to the artist whose model was used — attribution and compensation are built into the protocol rather than bolted on afterwards. Published works live at dedicated URLs with the full generation history visible: which tool, which artist's model, which prompt. Attribution, pricing, and settlement happen on Base (Coinbase's Ethereum L2).

The platform distinguishes artist-trained models (LoRAs created by named artists, which pay them royalties when used) from generic foundation models (no artist attribution). The Explore page and the Model Chooser in the Studio surface which models are artist-trained.

## Key pages

- [Vision](https://titles.xyz/notes/creative-infrastructure-that-remembers): What TITLES is building and why attribution is built into the protocol
- [Notes](https://titles.xyz/notes): Essays and updates from the TITLES team
- [Explore](https://titles.xyz/explore): Browse artist-trained AI models available on the platform
- [Feed](https://titles.xyz/feed): Community-published creations with full provenance
- [Studio](https://titles.xyz/create): Node-based creative workspace where generations happen (sign-in required, not crawlable)
- [Sitemap](https://titles.xyz/sitemap.xml): Machine-readable index of featured works, public models, and indexable creator profiles

## How attribution works

When a creator uses an artist-trained model to generate an image, a share of the generation fee automatically goes to that artist's wallet. Generations are pay-per-use and priced per-model — actual prices are shown in the Studio before each run. Artists earn royalties every time their model is used. Generic/foundation models don't carry artist attribution. All settlement is on-chain.

## Content types

Images: Text-to-Image, Image-to-Image, Image Edit, Image Variation, Blend Images, Image Transform. Three of these (Text-to-Image, Image-to-Image, Image Edit) support artist-model selection; the rest use fixed pipelines.

Videos: Text-to-Video and Image-to-Video. Both use fixed pipelines (no artist model selection today).

## Published works

Each published creation has a public page at `/published/{id}` showing the image or video at full size alongside the tool used, the artist's model, the prompt, and the creator. Non-featured published works are `noindex, follow` (shareable by direct link but excluded from search indexes). Only works the editorial team has featured appear in the sitemap and are indexed.

## Docs

Product documentation. Every page below is also available as plain markdown — append `.md` to its URL.

### Getting started

- [Getting Started with TITLES](https://titles.xyz/docs/getting-started-with-titles): What TITLES is, how artist-trained models work, and where to start.
- [Creating Images and Video](https://titles.xyz/docs/creating-images-and-video): From a blank canvas to a finished asset
- [Publishing and Remixing](https://titles.xyz/docs/publishing-and-remixing): Publish finished works to the community, share them anywhere, and let others build on them. Every contribution stays connected, so credit and royalties follow the work automatically.
- [Create with Your Assistant](https://titles.xyz/docs/create-with-your-assistant): Connect Claude, ChatGPT, Cursor, or another MCP-enabled assistant to TITLES, then create with artist-trained models without leaving the conversation.