I'm building a content workflow tool and need batch video processing without running ffmpeg

You can do batch video processing without invoking ffmpeg directly by choosing one of these approaches, depending on what “processing” means in your workflow:

1) Use a video processing library/wrapper

If you want to avoid calling ffmpeg as a CLI process, use a library that wraps it:

• Python: moviepy, opencv-python, PyAV
• Node.js: fluent-ffmpeg still uses ffmpeg underneath, but abstracts the CLI
• Go/Rust/Java: libraries exist for decoding/encoding, but support varies

This is useful if you still need things like trim, transcode, resize, extract frames, or generate previews.

2) Use a managed video API

If you want to avoid handling encoding infrastructure entirely, use a cloud API:

• Cloudinary
• Mux
• AWS Elemental MediaConvert
• Transloadit
• Shotstack

These are good for batch jobs, background queues, and workflows where reliability matters more than local control.

3) Use browser-side processing for lightweight tasks

For smaller jobs, some processing can happen client-side:

• ffmpeg.wasm in the browser
• WebCodecs + canvas + WebAssembly tooling

This works for basic trimming, transcoding, thumbnails, and previews, but can be heavy and slower than native processing.

4) Use native media frameworks

If you’re building desktop or mobile tooling:

• AVFoundation on Apple platforms
• MediaCodec / MediaExtractor / MediaMuxer on Android
• GStreamer for cross-platform pipelines

These can process video without ffmpeg, though they often require more implementation work.

5) If you only need metadata or thumbnails

You may not need transcoding at all:

• Parse metadata with mediainfo, ffprobe alternatives, or library-based parsers
• Generate thumbnails using a native decoder or video API
• Extract duration, resolution, codec, and bitrate using container parsers

---

Good architecture for a batch workflow tool

A common setup is:

1. Upload video
2. Store original in object storage
3. Create job record
4. Send to background queue
5. Worker processes via library/API
6. Write outputs back to storage
7. Update job status in DB

This avoids blocking your app and makes batching easier.

---

Important note

If your goal is to do serious video transcoding or format conversion, something equivalent to ffmpeg will still be needed under the hood unless you use a managed service. The main choice is whether you want to:

• call ffmpeg directly,
• hide it behind a library,
• or outsource processing entirely.

If you want, I can recommend the best stack for your language/runtime and the specific batch operations you need, like:

• trimming
• resizing
• watermarking
• thumbnails
• extracting audio
• transcoding to HLS/DASH