AutoShorts (gameplay) divyaprakash0426/autoshorts

Python tool that picks moments from long videos (gameplay first), crops to 9:16 and adds styled captions and optional AI voiceovers. NVIDIA GPU only.

Creator kit + Compare GitHub ↗ Install guide (official) ↗

AutoShorts (gameplay) at a glance

Checked 15 Oct 2026

AutoShorts is a Python CLI and Streamlit dashboard built for gameplay footage, with 15 content presets including podcasts and interviews. It picks moments with Gemini, OpenAI vision or local heuristics, transcribes with Whisper, adds PyCaps-styled captions and optional local Qwen3-TTS voiceovers. Apache-2.0. Needs an NVIDIA GPU with CUDA and NVENC; no CPU or Mac path. Crops with blurred backgrounds rather than speaker tracking. No releases, no publishing.

Cite this pageRepoQuarry. “AutoShorts (gameplay).” Checked 15 Oct 2026. https://repoquarry.dev/projects/autoshorts-gameplay

How much of Opus Clip does it replace?

3 of 5 jobs. Strongest on gameplay; cropping is not speaker tracking, and there is no virality score UI or social publishing.

AI picks highlight clips from a long video Partial replacement

Auto-reframe to vertical 9:16 (speaker or face tracking) Partial replacement

Animated burned-in captions on clips Partial replacement

Virality score per clip Not covered

Schedule and publish clips to social platforms Not covered

For virality score per clip, schedule and publish clips to social platforms, see other Opus Clip alternatives. You may also need to build an interface, an API server and hosting yourself.

Facts, with sources

Every answer links to where we found it. We read the documentation; we did not install it.

Report a correction

QuestionAnswerSourceChecked

Runs on my machine? Yes, local Python CLI and dashboard, or Docker with NVIDIA Container Toolkit `local` uses heuristic selection instead, but GPU rendering still requires NVIDIA hardware. github.com · autoshorts#get-started ↗ 1d ago

Code licence Apache-2.0 github.com · LICENSE ↗ 1d ago

Model weights licence Uses Whisper and optional Qwen3-TTS (Apache-2.0); weights keep their own licences huggingface.co · Qwen3-TTS-12Hz-1.7B-VoiceDesign ↗ 1d ago

Commercial use Yes, Apache-2.0 code; check the terms of any cloud AI provider you use Dependencies and model weights retain their own licenses. github.com · autoshorts#license-and-credits ↗ 1d ago

Paid services needed? Not required (local heuristic mode); Gemini or OpenAI analysis is optional and may cost money Gemini deep analysis uploads the full source video; API usage may incur charges. github.com · autoshorts#configuration ↗ 1d ago

Hardware NVIDIA GPU with CUDA and NVENC required; 6 GB+ VRAM and 16 GB RAM recommended for voiceovers For voiceovers, 6 GB+ VRAM and 16 GB system RAM are recommended. github.com · autoshorts#get-started ↗ 1d ago

Operating systems Not stated per OS; requires NVIDIA CUDA, so macOS is effectively excluded You need an NVIDIA GPU and driver, a CUDA toolkit with `nvcc`, FFmpeg with NVENC support, and Python 3.10+. github.com · autoshorts#get-started ↗ 1d ago

Setup effort Hard; make install builds CUDA-enabled Decord and needs the CUDA toolkit and FFmpeg 4.x libraries; Docker option Decord's build requires FFmpeg 4.x libraries github.com · autoshorts#get-started ↗ 1d ago