prime-skills

Generate, inpaint, and outpaint music using StepFun-AI's open-weights model with tag-driven composition and multilingual lyrics.

Create talking-head and lip-sync videos from audio using multiple AI avatar models optimized for different use cases.

Generate a custom Codex Pet spritesheet from a reference image using GPT image editing.

Generates images or videos conditioned on pose, skeleton, or motion references using multiple specialized models.

Generate full songs and instrumental tracks from text descriptions with section control, multilingual vocals, and studio-quality audio.

Swap faces or characters in videos and images, automatically selecting the best model for your content.

Edit images with Flux 1 Kontext Pro on RunComfy using optimized prompting patterns for precise local edits.

Generate images using Black Forest Labs' fast Flux 2 Klein model with optimized prompting patterns for higher quality output.

Generate and edit images using OpenAI's GPT Image 2 model with support for embedded text, logos, and multilingual typography.

Edit images using OpenAI's GPT Image 2 with optimized prompting patterns for precise results.

Generate videos from text using HappyHorse 1.0, featuring native 1080p output with synchronized audio and multi-language prompt support.

Route image editing tasks to the best specialized model for your needs—batch processing, text rewriting, local edits, or inpainting.

Remove objects, fill regions, and edit masked areas of images with mask-driven or description-based inpainting.

Extend images beyond their original canvas, change aspect ratios, and fill in surrounding content while preserving the original image.

Generate videos from text or images using Kling 3.0 with synchronized audio and consistent characters.

Synchronize mouth movements in video or avatars to audio tracks across multiple synthesis engines.

Generate images using Google's Nano Banana 2 model with optimized prompting patterns for sharper results.

Edit images while preserving subject identity, swapping backgrounds, and applying localized changes using spatial descriptions.

Change lighting setup, color temperature, direction, or mood in still images using specialized relighting models.

Generate and edit images using FLUX, Google Nano, OpenAI, and other models via intelligent model routing.

Generate and edit music using ElevenLabs or ACE Step models, with automatic routing to the most cost-effective option for your needs.

Generate videos from text or images using multiple AI models including Kling, Seedance, and Veo, automatically selecting the best model for your needs.

Run hundreds of AI models from the command line using the runcomfy binary for image generation, video creation, editing, and more.

Routes static images to the best RunComfy animation model for your needs—HappyHorse for general motion, Wan for lip-sync, Seedance for multi-modal sequences.

Routes video editing requests to the optimal model for restyles, background swaps, motion control, or outfit changes.

Generate cinematic short-form videos up to 15 seconds using ByteDance Seedance 2.0 Pro with multi-modal reference inputs.

Extend video clips by generating additional frames that continue the action, motion, and visual consistency of your original footage.

Remove objects, clean watermarks, or replace regions across video frames with motion-aware editing.

Extend video frames spatially and change aspect ratios while preserving the central action.

Generate videos from text prompts using Wan 2.7, with multi-reference conditioning and audio-driven lip-sync support.