I've compiled over 80 AI video editing tools

Yesterday I asked what tools you use to edit videos with AI.

Between your recommendations and what I've been looking into since, I've ended up compiling more than 80 tools, libraries and repositories.

There's a lot more out there than I expected.

The cool thing is that a lot of them can be combined with Claude Code to edit recordings, create animations, generate scenes, add voices or automate pretty much the whole process.

I've tried to organize them by what they're for. I've also included the links at the end 👇

1️⃣ Editing videos you've already recorded

There are a few ways to do this.

If you prefer a traditional editor:

🔸 Premiere: professional editor with AI features, editing from transcripts and a chat assistant in beta.
🔸 DaVinci Resolve: editing, color, effects and audio. It has a pretty full-featured free version.
🔸 CapCut: quick editing for social media, subtitles, templates and effects.
🔸 Descript: editing from a transcript, really useful for talking-head videos.

But you can also edit videos directly with code, using agents:

🔸 HyperFrames: HTML-based framework for putting videos together, adding effects, subtitles and animations. Includes templates and skills for agents.
🔸 Remotion: lets you create and edit videos using React. Really interesting for combining recordings with animated elements.
🔸 FFmpeg: cuts, transitions, zooms, color correction, audio, subtitles and rendering.
🔸 MoviePy: a Python alternative for automating editing tasks.

And there are tools that automate parts of the editing process directly:

🔸 Gling: removes silences, filler words and bad takes.
🔸 Eddie AI: lets you request edits via chat and export them to an editable timeline.
🔸 Kapwing: video editing using prompts and transcripts.

2️⃣ Transcripts, cuts and subtitles

I find this part especially interesting because it lets you edit the video based on what's being said.

🔸 Whisper / whisper.cpp: audio and video transcription, including locally.
🔸 WhisperX: transcription with word-level timestamps and speaker detection.
🔸 FFmpeg: silence detection and automatic cuts using scripts.
🔸 Auto-Editor: removes silences and automates cuts.
🔸 Captions: animated subtitles and automatic editing for short videos.
🔸 Submagic: subtitles, effects and visual elements for shorts.

For example, you can use WhisperX to get the transcript, ask Claude to decide which parts to remove, and run the cuts with FFmpeg.

3️⃣ Animations, motion graphics and infographics

This is where I've found the most interesting tools.

🔸 HyperFrames: animations, titles, graphics, overlays and compositions using HTML/CSS.
🔸 Remotion: motion graphics and animations coded with React.
🔸 GSAP: animations and precise timeline control.
🔸 Motion Canvas: explainer animations coded with TypeScript.
🔸 Manim: especially interesting for math and educational animations.
🔸 Lottie and Rive: reusable vector animations.
🔸 Three.js and React Three Fiber: 3D scenes and elements.
🔸 Blender: 3D modeling, animation and rendering.
🔸 After Effects: professional motion graphics, with different ways to automate them.

For animated graphics, product presentations or explainer videos, I'd start by checking out HyperFrames and Remotion.

4️⃣ Generating videos from scratch or modifying scenes

Things change quite a bit here. You don't necessarily need a recording anymore.

🔸 Kling: generates video from text or images.
🔸 Seedance: generates scenes with motion and audio.
🔸 Veo / Google Flow: video generation with scene direction tools.
🔸 Runway: video generation and editing existing footage.
🔸 Higgsfield: a platform that brings together models and tools for creating and transforming video.
🔸 Adobe Firefly: video generation integrated into the Adobe ecosystem.
🔸 ComfyUI: customizable generation workflows, including models you can run locally.

These tools generate or transform videos. Then you can use HyperFrames, Remotion, Premiere or FFmpeg to put all the scenes together.

5️⃣ Voices, music and sound

Audio is another part of the process you can automate quite a bit.

🔸 ElevenLabs: voice generation and transformation.
🔸 Kokoro: an open alternative for generating speech.
🔸 Chimera: combines ElevenLabs voice generation with Seedance video.
🔸 Sync: lip-syncs video with audio.
🔸 Adobe Podcast: improves and cleans up voice recordings.
🔸 Auphonic: automatic audio processing and normalization.
🔸 Instant Soundtracks: generates music and sound effects to go with videos.

You can also use FFmpeg to mix tracks, normalize volume or automatically lower the music when someone speaks.

6️⃣ Visual resources and inspiration

One thing a lot of you recommended was giving the agent visual references and components it can reuse.

Some interesting resources:
🔸 Prompt Motion: references and inspiration for animations.
🔸 60fps.design: examples of motion design and interactions.
🔸 Recraft: generates images and vector graphics.
🔸 Phosphor Icons: icons you can integrate into animations.
🔸 Backgrounds Supply: backgrounds for compositions.
🔸 Refero Styles: visual style references.
🔸 Figma MCP: lets you bring Figma designs and references into your workflow.

You can also give Claude references to animations you like and ask it to recreate them with HyperFrames, Remotion or GSAP.

7️⃣ Turning long videos into shorts and automating posts

Finally, there are more specialized tools for repurposing content and producing videos at scale.

🔸 OpusClip: turns long videos into short clips.
🔸 Riverside: podcast and interview recording, with tools for pulling out clips.
🔸 Creatomate: generates lots of video versions from templates and data.
🔸 Canva Bulk Create: creates visual assets in bulk.
🔸 Postiz: schedules and publishes social media content.

This could be useful for automating product videos, podcast clips or formats you repeat every week.

8️⃣ And what AI do I use to coordinate it all?

Looks like Opus 5.5 is king (someone also recommended Sonnet 5.5). But it seems like OpenAI's models aren't up to scratch.

The idea is for the agent to read transcripts, decide on cuts, generate animation code, run FFmpeg, render and review the result.

And here are a few repositories with editing skills and workflows already set up:
🔸 HyperFrames Skills
🔸 Remotion Agent Skills
🔸 ffmpeg-skill
🔸 video-use
🔸 Herk Edit Kit
🔸 Fragua

I think this is the part most worth exploring: how far we can go building a reusable editing workflow and controlling everything with prompts.

I haven't tried all these tools yet, so I wouldn't want to say which ones work best.

But between your recommendations and all the repositories I've been finding, there's plenty to experiment with for a while.

If you're using any combination that's giving you good results, share it. I'm especially interested in what you're managing to do with Claude Code, HyperFrames, Remotion and FFmpeg.

Editing:
capcut.com/
descript.com/
blackmagicdesign.com/products/davincires…
adobe.com/products/premiere.html

Clips:
opus.pro/
riverside.com/

Agents:
palmier.io/
github.com/ggml-org/whisper.cpp
github.com/m-bain/whisperX
github.com/kajisho5/ffmpeg-skill
auto-editor.com/

Animation:
hyperframes.heygen.com/
remotion.dev/
adobe.com/products/aftereffects.html

Generation and transformation:
kling.ai/
seed.bytedance.com/en/
flow.google/
higgsfield.ai/
firefly.adobe.com/
runwayml.com/

Audio:
elevenlabs.io/
podcast.adobe.com/en/enhancespeech
auphonic.com/