+ 
+ 
++ A premium, zero-configuration local AI studio and offline GUI for Stable Diffusion (Image Generation), LLMs (Chat), Whisper (Speech-to-Text), and Kokoro (Text-to-Speech). Powered by hardware-accelerated GPU and NPU execution on Windows, Linux, and macOS. +
+
+
+ π₯ Watch the Setup & Demo Video: https://youtu.be/ESELhY-G_9w +
-| **Generation Workspace** | **Model Library** | **Image Constraints** | -| :---: | :---: | :---: | -|
| 
| 
|
-
----
-
-
-
-π₯ Watch the Setup & Demo Video: https://youtu.be/ESELhY-G_9w
+
- 
+
-
Run scripts/reset/reset.ps1 (Windows) or scripts/reset/reset.sh (Linux/macOS). This will clear temporary compilation and package caches to repair your environment. (Note: This preserves your model weights and generated output images).
GLIBC_2.38 not foundThe prebuilt binaries require glibc 2.38+ (e.g. Ubuntu 24.04). If your distribution uses an older glibc version, you can upgrade your operating system or compile the backend from source (see the Building From Source guide below).
+The web user interface runs on port 1420 by default. The GPU backend manager attempts to bind to port 8080 first, then automatically detects and falls back to a free system port if 8080 is already occupied.
Ensure your AMD GPU hardware and host kernel are fully compatible with ROCm 7.13. If ROCm fails to initialize correctly, the application will automatically fall back to Vulkan acceleration.
+3221225781 (0xC0000135)This code means Windows could not locate a required backend DLL:
+app/backend/win/vulkan/.This indicates that the local backend engine process terminated. Check your launch terminal (where you executed windows.bat, ./linux.sh, or ./mac.sh) for the exact console error. Common causes include glibc version mismatches, missing Vulkan drivers, or system out-of-memory (OOM) issues.
- {health?.stale - ? `Running server build: ${health.build || "unknown"}` - : readinessIssues.length > 0 - ? "Local AI Image Generator found setup items that may need attention." - : `${cleanupItems.length} temporary item${cleanupItems.length === 1 ? "" : "s"} can be cleaned (${formatBytes(cleanupBytes)}).`} -
-- Configure size, quality, and performance controls for image generation. -
-{activeModel}). Generating at larger sizes (like 1024x1024) is disabled to prevent out-of-memory crashes on your {specs && specs.gpu_name && !specs.gpu_name.includes("Loading") ? specs.gpu_name : "graphics processor"}.
- {specs.gpu_name}) is a GTX-series GPU which lacks hardware Tensor Cores. Running in CUDA mode will be up to 3x slower. We strongly recommend switching to Vulkan GPU for optimal generation speed.
- + The local C++ llama.cpp server is running. Chat requests can be sent via the Text Chat interface. +
+ {backendInfo.llmBackendMode && ( ++ The local whisper.cpp runtime is ready. Speech Transcriber can process WAV recordings and uploads. +
+ {backendInfo.speechBackendMode && ( ++ The local Kokoro ONNX runtime is ready. Text to Speech can generate WAV narration. +
+ {backendInfo.ttsBackendMode && ( +- No models detected in app/models/. Download from the library below or import a file. + {activeModelType === "image" + ? "No image models detected in app/models/. Download from the library below or import a file." + : activeModelType === "text" + ? "No text models detected in app/llm-models/. Download from the library below or import a file." + : activeModelType === "speech" + ? "No speech models detected in app/speech-models/. Download from the library below or import a .bin file." + : "No TTS models detected in app/tts-models/. Install Kokoro from the library below or import a .json manifest."}
) : (
{model.notes}
@@ -698,10 +1753,10 @@ function ModelManager({ activeModel, setActiveModel, serverRunning, setServerRun
disabled={installed || downloadingModelId !== null}
>
{installed ?
- Download any GGUF or Safetensors model from Hugging Face or other sites directly to your models folder. + Download any {activeModelType === "image" ? "Safetensors" : activeModelType === "text" ? "GGUF" : activeModelType === "speech" ? "Whisper .bin" : "Kokoro manifest"} model from Hugging Face directly to your models folder.
{isUrlDownloading ? (- "{activeModel}" is already loaded. Unload it before loading "{pendingLoadModel}". + "{pendingLoadModel.activeRuntime?.model}" is already loaded as a {pendingLoadModel.activeRuntime?.label || "model"} model. Unload it before loading "{pendingLoadModel.modelId}" as a {getModelTypeLabel(pendingLoadModel.targetType)} model.
+ The model {vramWarning.modelId} is approximately {vramWarning.modelSizeGb.toFixed(1)} GB, which exceeds your currently available GPU VRAM of {vramWarning.freeVramGb.toFixed(1)} GB free (out of {vramWarning.totalVramGb.toFixed(1)} GB total).
+
+ Loading this model on your GPU may fail, crash the backend, or cause extremely slow generation due to memory spilling.
+