One-Click AI Tool Packs

Portrait Background Removal

Rembg removes image backgrounds automatically. Use it as a CLI, Python library, HTTP server, or Docker container.

Document OCR & Layout

Converts documents and images into structured, AI-friendly data such as JSON and Markdown.

Real-time Digital Human

Real-time talking-head generation from audio and a single portrait image, developed by Soul AI Lab.

Text-to-Speech

Open-source TTS from Alibaba Qwen team. Supports expressive streaming speech, voice design, and voice cloning.

Lip Sync for Digital Humans

Generate lip-synced talking-head videos from audio and reference portraits.

Text-to-Video

Generate short videos from text prompts with the LTX 2.3 text-to-video pipeline.

Inpainting & Watermark Removal

Remove watermarks, fill missing regions, and erase objects from images with inpainting models.

Motion Transfer

Transfer motion from a reference video to a still image and generate animated results.

Vision-Language OCR

Accurate layout parsing for skewed, scanned, low-light, and phone-captured documents.

Image/PDF to Markdown

Structured document parsing framework that turns images and PDFs into pixel-accurate Markdown output.

Virtual Try-On

One-click virtual outfit try-on for fashion and e-commerce workflows.

Video Background Removal

Remove backgrounds from videos with high-quality matting for compositing and editing.

Multi-Angle Face Edit

Edit faces from multiple angles with Qwen image editing models.

Face Swap

Industry-popular face swapping toolkit with CUDA acceleration for local GPU inference.

Head Swap

Lightweight head-swap workflow built on FLUX.2 Klein for fast local generation.

Speech to Subtitles

Fast Whisper-based transcription for turning speech into timed subtitles locally.

Aligned Transcription

Whisper transcription with word-level timestamps and speaker diarization support.

Speech Recognition

Automatic speech recognition pack for converting audio into text and subtitles.

Garment Extraction

Extract clothing items from photos for virtual try-on and catalog workflows.

Image Segmentation

Segment Anything Model for interactive and automatic object segmentation in images.

Streaming Transcription

Dockerized streaming ASR for real-time speech-to-text on local hardware.

Open-source Digital Human

Open-source AI avatar pipeline for creating talking digital humans locally.

Edge Speech Toolkit

All-in-one edge speech stack: ASR, TTS, VAD, and speaker ID with ONNX runtime.

Offline Speech Synthesis

Fast, lightweight offline text-to-speech engine with many voice models.

Offline Speech Recognition

Open-source offline speech recognition toolkit for 20+ languages on CPU.

Lightweight Offline TTS

Compact ONNX-based TTS model for low-resource offline speech synthesis.

CPU-friendly OCR

Lightweight OCR engine that runs well on CPU without a dedicated GPU.

PDF to Markdown

High-quality PDF and document conversion to clean Markdown with layout preservation.

Image & Video Upscale

Classic open-source super-resolution for enhancing image and video quality.

Face Restoration

Restore and enhance faces in old or low-quality photos and videos.

Photo Colorization

Colorize black-and-white photos and footage with deep learning models.

Classic OCR Engine

The classic open-source OCR engine supporting 100+ languages offline.

Lightweight Image Generation

Compact 4B-parameter FLUX.2 Klein model for fast local AI image generation.

PDF Parsing

Open-source PDF parsing toolkit for extracting structured content from documents.