LoRA Pilot is an ultimate docker image for all Stable Diffusion LoRA trainers. Includes kohya_ss, diffusion pipes and TensorBoard for trainings and ComfyUI and InvokeAI for validation. Features shared models, modules, custom integrations and automatization scripts.

Production-ready RunPod serverless endpoint and pod for Qwen-Image (20B) - Text-to-image generation with exceptional English and Chinese text rendering

Production-ready RunPod serverless endpoint for Kokoro TTS. Features high-quality text-to-speech, voice mixing, word-level timestamps, and phoneme generation. Optimized for fast cold starts and auto-scaling.

An all-in-one docker image that runs the latest ComfyUI with SageAttention.

Serverless GPU Stable Diffusion XL worker for Runpod, powered by diffusers. Generate images from text, transform existing images, or inpaint masked regions — all from a single serverless endpoint. Supports SDXL Base 1.0, SDXL Turbo, SDXL Lightning, Juggernaut XL, Playground V2, plus the official SDXL refiner for a final detail pass.

Serverless GPU/CPU zero-shot text classification on RunPod. Hand the worker any text and an arbitrary list of candidate labels, get back ranked probabilities in milliseconds — no fine-tuning, no training data, no label schema baked into the model. Powered by NLI-style encoders such as BART-MNLI, DeBERTa-v3-MNLI, and multilingual checkpoints...

Serverless, pure-CPU PDF processing worker for Runpod. Extracts text, layout-preserving text, Markdown, tables, metadata, embedded images, hyperlinks, bookmarks/outline, form-field values, redaction/annotation listings, and page renders (PNG/JPG). Also runs in-document text search and produces concatenated output. Backed by PyMuPDF (fitz) and ...

Serverless CPU-only MediaPipe Tasks worker covering the full vision catalog through a single endpoint: pose estimation, hand landmarks, face mesh, holistic (pose + hands + face), object detection, image segmentation, and gesture recognition — plus an annotation mode that overlays everything onto the input image.

A RunPod serverless worker for Microsoft TrOCR, a HuggingFace transformer-based OCR model specialized for single-line text recognition — both printed and handwritten.

Serverless GPU LLM inference worker built on vLLM with an OpenAI-compatible request shape. One worker, one loaded model — vLLM owns CUDA memory and refuses model swaps at runtime, so picking your backbone is an env-var concern, not a per-request one. Switch via LLM_MODEL and redeploy.

Serverless GPU worker that turns a text prompt into a short video clip using AnimateDiff — a research-grade motion module that bolts onto any Stable Diffusion 1.5 or SDXL checkpoint and gives it temporal awareness. Served via the diffusers library on Runpod serverless. Pick a base model, pick a motion adapter, optionally stack a camera-motion LoR

Serverless speech-enhancement worker for RunPod, built on DeepFilterNet3 — a real-time deep-learning noise suppressor that consistently tops the DNS-Challenge / VCTK / VoiceBank-DEMAND benchmarks while running CPU-fast.

A serverless OCR worker for RunPod powered by PaddleOCR and PP-Structure, providing high-accuracy multilingual text recognition and structured document understanding. It delivers stronger performance for Asian scripts (Chinese, Japanese, Korean), Arabic-script languages (Arabic, Persian/Farsi, Urdu), and complex document layouts including tables...

A serverless multi-task NLP toolbox built for scalable deployment on RunPod. Expose a single API endpoint to handle multiple natural language processing tasks with low overhead and production-ready performance.

A RunPod serverless worker for Donut (Document Understanding Transformer), NAVER's OCR-free document AI model. Donut takes a document image plus a task prompt and produces structured JSON directly — no separate OCR step, no detector pipeline. The model is small (~200M params), fast on a single GPU, and remarkably accurate on layout-driven documents

Serverless background removal worker powered by rembg on Runpod. Accepts public image URLs or base64-encoded images and returns the cut-out RGBA PNG (or WebP/JPG), an optional alpha mask, an optional composited image on a custom background color, and the foreground bounding box. Supports the U2Net, IS-Net, SAM, and BiRefNet model families with GPU

Serverless GPU worker that turns a still image into a short video clip using Stability AI's Stable Video Diffusion family of checkpoints, served via the diffusers library on Runpod serverless. Send a PNG/JPG (URL or base64), get back an mp4, an animated GIF, or the raw frame list.

Runpod Serverless GPU image super-resolution powered by Real-ESRGAN with optional GFPGAN face restoration. Accepts image URLs or base64 inputs, lazy-downloads model weights to disk on first use, supports tile mode for huge images, and returns base64-encoded PNG / JPG / WebP outputs.

Serverless GPU endpoint for Meta's Segment Anything 2 (SAM 2). Supports bounding-box prompts, point prompts, multi-prompt batches, an automatic "segment everything" mask generator, mask refinement, and multi-image batches. Optional Grounded-SAM-style text prompts via GroundingDINO are wired in when text_prompts are provided.

A serverless GPU text-to-speech worker for RunPod that combines multiple speech synthesis engines under a single API endpoint. It supports Coqui XTTS-v2 for high-quality multilingual speech generation and zero-shot voice cloning from short reference audio clips, Coqui standard TTS models including Tacotron2, Glow-TTS, and VITS for single-language..