ATH-MaaS/Pixelle-Video

Health v2 · maintenance only

Adoption not part of the score

28735 stars · 4178 forks observed · 2026-10-08

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

Pixelle-Video is an AI-powered fully automated short video engine that turns a single topic into a finished video. It automatically writes the script, generates AI images and video clips, synthesizes voiceover via TTS, adds background music, and composites the final video through a modular pipeline built on ComfyUI/RunningHub workflows or direct model APIs.

Use cases

• generate short videos from a text topic automatically
• create faceless youtube or tiktok videos with ai narration
• automatically write video scripts with an llm
• generate ai images and video clips for each scene
• add text-to-speech voiceover and background music to videos
• create digital human talking-head videos
• batch produce short videos without editing experience

When to choose

• you want one-topic-to-finished-video automation with zero editing skills
• you want to swap in different LLM, image, video, and TTS providers via ComfyUI workflows or direct APIs
• you need batch video creation with a web UI and multiple templates and aspect ratios

When to avoid

• you need fine-grained manual control over every edit in a traditional video editor
• you cannot access external AI model APIs or GPU-backed services like ComfyUI/RunningHub
• you need long-form video production rather than short-form clips

Facets

application · maturity active

video-processing tts image-processing llm-inference stable-diffusion workflow-automation web-framework artificial-intelligence media large-language-models python cross-platform windows aigc short-video-generation comfyui text-to-video ai-narration digital-human faceless-video video automation docker web-server

Member repositories

For agents

markdown · JSON · MCP: product_card(name="ATH-MaaS/Pixelle-Video")