A downloadable tool



UPDATE:

AI Runner 3.0 is right around the corner. In the meantime I have disabled downloads of AI Runner. Your existing copy will still work, and you can still run the old version by downloading the code from Github.

The upcoming version has improved workflows and major new features such as the ability to run LLMs for chatbots, prompt generation, voice chat and more.

Bookmark this page and follow along on the github product page while you await the upcoming 3.0 update.

Run Stable Diffusion and Kandinsky on your own hardware without a webserver or additional requirements to install

With AI Runner you are able to run Stable Diffusion, Kandinsky and other diffusion AI models on your own computer without any technical knowledge. Just download the app, pair it with your favorite model and run.

Drawing tools

A simple but growing set of art tools paired with image-to-image allow you to convert sketches into works of art.

Sketch of a mountain

Before applying image-to-image

Stable Diffusion generated image of a mountain using original sketch

After applying image-to-image with the prompt "A Painting of a Mountain"


Infinite canvas with collapsible panels

An infinite canvas allows you to outpaint images to any size and collapsible panels make it easy to view your work on a clean interface as you do so.

AI Runner with panels collapsed


Image filters

Modify your creations with image filters, such as the pixel art filter and film filter to give them a unique look and feel.



Customizable

Choose custom paths to store your models, embedding, LoRAs, images, videos and gifs. Tweak memory settings and more.


Easy prompt storage

Save your favorite prompts, edit and reload them with the prompt browser.


Features


Category Features
Image Generation Generate images using Stable Diffusion and Kandinsky models
Flexible prompts and weighting using compel and automatic1111 style prompt weighting
LoRA and Textual Inversion / Embeddings
Support for various model types and formats such as diffusers, safetensors, and ckpt
Customizable metadata import / export options
Automatic image export
Control grid size, line width, and color
Drawing tools for transforming sketches into art
Infinite scrolling canvas
Outpainting for creating large-scale artwork
Image batch generation and interpolation
Image filters for creative effects
Usability and Performance Easy setup and installation
Fast image generation with GPU acceleration**
CPU image generation†
Memory preferences for optimized GPU RAM usage
Fast load time and responsive interface
Additional Features NSFW filter toggle
Stored sessions - all of your settings and prompts for each section are stored and automatically loaded between sessions.
Active development with regular updates
Architecture Torch 2 and memory enhancement settings
No additional installation requirements, and run*
* After running the program, you will need to download the stable diffusion models. This requires an active internet connection. Optional copy+paste requires xclip on Linux.
** Image generation results are dependent on hardware.
CPU support is limited, and we do not recommend running AI Runner without a CUDA capable GPU.

Model pipelines and tools


Stable Diffusion
Kandinsky Zeroscope Shap-e
txt2img
img2img
depth2img
txt2vid
Image interpolation
pix2pix
inpaint
outpaint
controlnet
upscale
superresolution
Compel prompt weighting
LoRA
Textual inversion / embedding
Deterministic image batches
Prompt builder
Prompt saver
Model Merger

GPU Memory options

Accelerated Transformers
Attention Slicing
Channels last memory
Sequential CPU offload
Model CPU offload
TF32
Vae slicing
Tile vae

Minimum System Requirements

Cuda capable GPU (2080s or higher recommended)
8gb of system RAM (16gb recommended)
6gb of VRAM (8gb+ recommended in order to run all features)*
5.8gb of disc space to install AI Runner (100gb+ recommended for custom models)**
In order to reliably use AI Runner on systems with low VRAM we recommend closing all other applications.

* May run on as little as 3gb VRAM with correct memory settings, however this is untested.

** The core AI Runner  program takes approximately 5.8gb of disc space to install, however the size of each model varies. Typically models are between 2.5gb to 10gb in size. The more models you download, the more disc space you will need.


AI Runner is an early access program built with open source software. Absolutely no guarantee or license is offered with this application.

Depending on the model used, AI Runner may be able to generate


The following use restrictions apply to Stable Diffusion models and are part of the Stable Diffusion license

Use Restrictions

You agree not to use the Model or Derivatives of the Model: 

  • In any way that violates any applicable national, federal, state, local or international law or regulation
  • For the purpose of exploiting, harming or attempting to exploit or harm minors in any way; 
  • To generate or disseminate verifiably false information and/or content with the purpose of harming others;
  • To generate or disseminate personal identifiable information that can be used to harm an individual;
  • To defame, disparage or otherwise harass others;
  • For fully automated decision making that adversely impacts an individual’s legal rights or otherwise creates or modifies a binding, enforceable obligation;
  • For any use intended to or which has the effect of discriminating against or harming individuals or groups based on online or offline social behavior or known or predicted personal or personality characteristics;
  • To exploit any of the vulnerabilities of a specific group of persons based on their age, social, physical or mental characteristics, in order to materially distort the behavior of a person pertaining to that group in a manner that causes or is likely to cause that person or another person physical or psychological harm;
  • For any use intended to or which has the effect of discriminating against individuals or groups based on legally protected characteristics or categories;
  • To provide medical advice and medical results interpretation;
  • To generate or disseminate information for the purpose to be used for administration of justice, law enforcement, immigration or asylum processes, such as predicting an individual will commit fraud/crime commitment (e.g. by text profiling, drawing causal relationships between assertions made in documents, indiscriminate and arbitrarily-targeted use).

THIS SOFTWARE IS PROVIDED 'AS IS', WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

BY PROCEEDING WITH THE DOWNLOAD AND INSTALLATION OF THIS SOFTWARE, YOU ACKNOWLEDGE AND CONSENT TO THE AFOREMENTIONED TERMS.



Updated 26 days ago
StatusIn development
CategoryTool
Rating
Rated 4.1 out of 5 stars
(9 total ratings)
AuthorCapsize Games
Tagsai, art, artificial-intelligence, Asset Pack, Generator, Meme, Pixel Art, stable-diffusion, tool, User Interface (UI)
Average sessionDays or more
LanguagesEnglish
InputsKeyboard, Mouse

Development log

View all posts