Docs / Demo projects
Text to image
Open text-to-image models turn one prompt into any number of pictures on your worker, shown live as they are made, and compared from small to large.
assets/images/site/demo-text-to-image.jpg1600 × 900 pxThe Images view filling with pictures of a rover in a red Martian canyon, one per seed, with a chart comparing the seconds per image of four models.Open text-to-image models from Hugging Face, downloaded once onto your worker, turn a prompt into pictures. Each one appears on the View page as it is made.
The flows
| Flow | What it does |
|---|---|
| 1 · One prompt, any number of images | Segmind Vega turns one prompt into 20 images (seeds 1 to 20), each shown live. |
| 2 · One prompt, ten styles | The same scene in ten styles (sunrise, night, dust storm, watercolour, pencil sketch…), two seeds each. |
| 3 · Small to large: four models | The same prompt and seeds on Vega RT, Vega, SSD-1B and SDXL 1.0: their images side by side and the time each takes. |
What you need
An NVIDIA GPU, a Tenstorrent card or the CPU (slower, with the smallest model). The first run downloads the model (about 3.3 GB for Vega). The worker needs Run code from projects turned on (Computing > your worker > Worker settings).
Try it
- Open Projects > Demos and choose Text to image. Read its readme.
- Open flow 1 in the Blueprint and press Run: it runs as it is on your own workers. Switch to the View page to watch it. Its runs and results are yours only.
- To change it, choose Clone into my workspace: a copy goes into your workspace, with its readme, blueprints, the code of its blocks and its files, saved to a GitHub repository of yours (link GitHub first, under Profile > GitHub).