AIAny
AI Video2025
Icon for item

Y2A-Auto

Automatically transfers YouTube videos to AcFun and bilibili with an end-to-end pipeline: downloading, ASR, subtitle translation and QC, AI-generated metadata, content moderation, and automated uploads; includes a web dashboard and monitoring.

Introduction

This project addresses a common bottleneck for cross-posting video content at scale: stitching together downloading, speech-to-text, subtitle translation and QC, AI metadata generation, content moderation, and platform uploads into a single maintainable pipeline. It shifts manual reposting tasks into configurable automation while keeping human-in-the-loop controls for review and safety.

What Sets It Apart
  • Integrated AI steps so you can run ASR (Whisper/Voxtral), subtitle translation, AI-based subtitle QC, and AI-generated titles/descriptions/tags in one flow — so what: reduces repetitive manual editing and speeds up multilingual reposting.
  • Platform-specific automation: built-in AcFun/bilibili upload modules, QR/Cookie login support and CookieCloud sync — so what: lowers the friction of platform authentication and supports scheduled monitoring of channels/keywords.
  • Deployment-friendly design: Docker-first with optional local mode, hardware-accelerated encoding (HEVC fallback to H.264), and configurable concurrency — so what: fits small self-hosted servers to GPU-enabled instances without heavy rewrites.
  • Safety and observability: password protection, brute-force lockout, content-moderation hooks (Aliyun Green), and async notification channels — so what: makes unattended automation safer and auditable.
Who It's For and Tradeoffs

Great fit if you need to repeatedly repost multilingual YouTube content to Chinese video platforms and want automation for subtitles and metadata while preserving manual audit gates. Not ideal if you require strict platform compliance/legal guarantees from the upstream content owner or need a zero-dependency hosted SaaS; it’s self-hosted, requires valid cookies/API keys, and may need tuning for large-volume or highly regulated pipelines.

More Items

Hugging Face
AI Video2026

Distills MiniMax H3 video generation into 2 or 4 steps, with an option to predict pixels directly and skip VAE decoding. Supports synchronized audio and first- or last-frame conditioning.

Hugging Face
AI Model2026

Generates video with audio from prompts using a MiniMax H3 fine-tune, with an uncensored tuning approach. Optional quantized checkpoints reduce memory demands; the weights are large and require local inference.

GitHub
AI Video2026

Generates MiniMax H3 videos locally on consumer hardware, adapting its inference path to available memory and compute. It can run with as little as 8 GB of VRAM and 16 GB of RAM.