OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Image & Video AI

Mine StableDiffusion

Mine StableDiffusion is a Kotlin Multiplatform app for generating images with Stable Diffusion models locally on Android, iOS and desktop. It pairs a Compose Multiplatform interface with a native C++ inference backend based on stable-diffusion.cpp.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Mine StableDiffusion — OpenViST project profile in Image & Video AI.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when you want to run image-generation models locally and your device can handle the chosen model. The README recommends trying SD-Turbo or SD 1.5 to gauge device capability before moving to more demanding architectures.

Who it is for

  • People who want to generate prompt-based images on a supported mobile or desktop device.
  • Creators who want to compare model families, sampling methods or LoRA combinations.
  • Developers exploring a Kotlin Multiplatform image-generation app with a C++ inference backend.

Documented uses

  • Generate images from text prompts, including batches of up to 10 iterations.
  • Use supported Stable Diffusion model families, with model choice guided by device resources.
  • Add and combine .safetensors LoRAs and adjust their weights.
  • Switch sampling algorithms and export PNGs that retain generation settings.
Before you choose

Requirements & limits

  • Model requirements vary: the README calls for ample VRAM and modern GPUs for high-detail generation with professional-tier models.
  • The documented stack includes Kotlin Multiplatform, Compose Multiplatform, a C++ inference backend, JNI, Vulkan and Metal.
  • CPU offloading and CPU placement options for text-encoder or VAE work are documented for selected memory or crash scenarios.
  • The public repository reports a GPL-3.0 license.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Generate images with a local Stable Diffusion model

Use Mine StableDiffusion to create images on a supported device with a model that fits the device’s memory limits.

Before you start

  • Mine StableDiffusion on Android, iOS, or desktop.
  • A supported Stable Diffusion model suited to the device’s available VRAM; the project recommends starting with SD-Turbo or SD 1.5 to gauge device capability.

Steps

  1. Choose a model appropriate for the device’s VRAM; start with SD-Turbo or SD 1.5 if assessing capability.

  2. Use the app’s text-to-image generation feature to generate an image from a prompt.

  3. If model loading runs out of memory, enable Offload to CPU; if generation crashes, enable Keep CLIP on CPU; if decoding runs out of memory, enable Keep VAE on CPU.

Expected outcome

The app generates images locally from text prompts; exported PNGs include generation parameters such as prompt, seed, model, sampler, CFG, and LoRAs.

Guide limitations

  • Model requirements vary; higher-quality architectures may require modern GPUs with ample VRAM.
  • CPU offloading can address memory issues, but the source does not document its performance impact.
  • The source describes supported platforms and features but does not provide a complete platform-specific operating or installation procedure.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
16 October 2026
Original website title
Onion99/KMP-MineStableDiffusion · Official homepage title · retained in its source language
How we review sourcesSuggest a correction