generative-video

A virtual camera films an AI agent's screen while a tracker locks onto every glyph. Two fframes passes, one SkSL shader, all sound synthesized.

The open .char format for consistent characters. Face, body, wardrobe & voice in one file, consistent across every model.

AI-native video editing built around real creative timelines, generative workflows, and local agent control.

Open-source local implementation of MiniMax H3 Context-IR. Turn simple prompts into structured, validated H3 video briefs.

Open Film Skills: 21 independent modules for story, directing, visual assets and AI-film production, with research galleries and playable 3D previs.

Open-source Next.js SaaS for Seedance 2.0 , Seedance 2.5 and Seedance 2 Mini video generation — Stripe billing, credits, NextAuth, and Prisma out of the box.

A curated collection of AI video prompting resources, featuring official guides, prompt templates, cinematic techniques, and audio-visual synchronization methods for Veo, Sora, Runway, Pika, Kling, and other leading AI video tools.

Realtime interruptible science tutoring with continuously generated H3 Max video.

Hermes Agent skill for typography-led lyric videos using GPT Image 2, Gemini Omni Flash, and exact-text QC

Four ComfyUI nodes for MiniMax H3: name your media with @, lock exact dialogue, and get the structured brief H3 wants. No <Picture 1> bookkeeping, no second service to start.

VS Code-inspired rich H3 prompt editor and reference palette for ComfyUI

334 verified prompts people used to make videos with Claude Opus 5.5, copied word for word from each creator's post. Previews, EN/中文 translations, and a director's toolkit.

Open-source Veo video generator — production-ready Next.js SaaS for text-to-video, image-to-video, and multi-image editing powered by Google Veo. Stripe billing, credits, NextAuth, and Prisma out of the box.

Beyond Time Shifts (ECCV 2026): a reference-free audio-visual synchronization evaluator built on Qwen2.5-Omni-3B, with SynthSync data, R-GRPO training, and the SyncBench benchmark.

Advanced, end-to-end, enterprise-grade agentic AI pipeline that automates competitor ad intelligence, performs multimodal creative strategy extraction, enables brand-safe adaptation, and generates AI video ads using LLM reasoning, multimodal analysis, and deterministic workflow orchestration with full auditability.

Agent skill + template: short videos made entirely from code. One HTML file, every frame a pure function of (frame, seed, width). Claude Code, Codex, Gemini CLI, Cursor.

A structured, JSON-based schema for precise, multi-scene AI video prompting and cinematic control.

Turn a story into a finished film — on your own Mac, no cloud. Story Forge is a local-only generative cinema pipeline: stills, motion, voices, music, grade, and sound, start to finish. Flux · LTX-2 · Piper · ACE-Step · ffmpeg — built for and running 100% on a Mac (M5, 128GB).

Motion graphics as code. An agent skill for building frame-exact animated logos, kinetic typography and brand videos as a single HTML file, then exporting them to MP4 in 1:1, 4:5 and 9:16.

Build an interactive, cinematic choose-your-own-adventure world you click through, from nothing but a concept or from your photos, with your own coding agent (Claude Code, Codex or Hermes), the Sogni Creative Agent Skill and a Sogni API key. Nine open-source models paint, film, narrate, score and build it, in one overnight session.