Free open-source Chinese AI tool turns one idea into a finished video, 110K GitHub stars
Shruti_0810 · x · 2026-09-07
An open-source Chinese AI video tool with 110K+ GitHub stars turns a single idea into a finished video: it writes the script, matches keywords to stock footage from Pexels, Pixabay and Coverr, adds voiceover, subtitles and music, supports landscape/vertical formats, and can publish directly to TikTok, Instagram and YouTube — up to 10 versions from one idea. No powerful PC or voice API key needed; usable via web, API, CLI, or fully autonomously by an AI agent.
Related event: Open-source MoneyPrinterTurbo turns a single idea into fully edited videos(2 posts)→
More from Multimodal
- A shark-as-villain clip made with Seedance 2.5 shows off an original prompt — techhalla · 2026-09-07
- Snorricam H3 LoRA released on HuggingFace and Civitai for extreme POV shots — NebSH83 · 2026-09-07
- GPT-6 Astra makes a kids' cartoon with Higgsfield, edits in Premiere, uploads to YouTube — LexiLove · 2026-09-07
- Minimax-h3 Fine-tuned Video Model Trends on Hugging Face — WarmBloodAban · 2026-09-07
- Magnific releases 2-minute AI animated short 'Slice' with all prompts published — cuenca · 2026-09-07
- Open-Source AI Movie Studio UI Adds LoRA, Model, and API Controls — purecharisma2020 · 2026-09-07