by
Georgie Kemp

Introducing OpenEdit

No items found.

What's new in OpenEdit?

Updated 19 August 2026

Two additions went live this week:

Fabric integration. OpenEdit can now generate its own assets rather than waiting for you to supply them. It can generate footage with VEED Fabric through the VEED API, produce images, music and sound effects in whichever model suits the job, build a soundtrack from several tracks at once, and handle longer videos than before. Source video was always optional in principle, and with Fabric wired in it holds up in practice: stills, slides, generated images or pure motion graphics are enough on their own. Generating anything spends credits, and OpenEdit asks first.

Motion and editing in the skill. Cutting, reframing and layering motion graphics and visual elements now sit in the skill itself, so you can ask for them directly instead of coaxing them out. This path is newer than captioning, and rougher.

Coming soon: Windows support, Lip Sync V2 and background removal.

OpenEdit is an open-source, agent-driven editing pipeline that ships with VEED's HTML renderer (which is closed source, but free to use)

There is no GUI and no timeline. The pipeline is driven entirely through your coding agent. Edit, cut, and reframe footage. Layer motion graphics and visual elements. Turn slides or websites into video. You can capture web pages, pull in any video or image generating service, or any MCP server when it helps. Source video is optional: stills, slides, generated media, or pure motion graphics are enough when the brief calls for it.

You compose in HTML and CSS, and VEED's renderer turns that into video on your own machine, with no headless browser in the pipeline. The OpenEdit pipeline is open source under Apache-2.0. It's free to use: no subscription, no paywalls.

OpenEdit is available now on GitHub.

Quick facts

  • What it does: agent-driven editing pipeline: cut and reframe footage, layer motion graphics, turn slides or websites into video, capture web pages, add captions, generate footage and audio when you have none
  • Install: npx skills add veedstudio/open-edit
  • Agents: agent-agnostic. Tested with Claude Code, Codex, and Gemini
  • Time to result: 1 to 3 minutes for a short clip, prompt to a finished MP4. Longer videos take longer
  • Renderer: VEED Engine, an HTML-to-video renderer that runs locally, with no headless browser in the pipeline
  • Speed: up to 2.2x faster than tools that drive Chrome, per VEED's early internal benchmarks
  • Price: free to use. No subscription, no paywalls
  • Licenses: the OpenEdit pipeline is Apache-2.0. Only VEED's renderer binaries are closed source, under PolyForm Shield 1.0.0, and free to use
  • Account: one free veed.io account for transcription. Transcription is built in. The default workflow gives you access to VEED's transcription when you provide a VEED account. Your audio — not the video — is uploaded and stored to transcribe it, and the limits are whatever your VEED account already has.
  • Requirements: Apple Silicon Mac on macOS 26 (Tahoe). Not Intel, not earlier macOS. Windows support arriving soon
  • Available now: github.com/veedstudio/open-edit

What is OpenEdit?

OpenEdit is two things: an open-source editing pipeline your AI agent drives and the HTML renderer it renders with. You run one install command, open your agent, and describe the job. The pipeline handles composition, and the renderer turns it into video on your machine.

Composition happens in HTML and CSS, which is the open-source part. The renderer then turns that composition into video. Because composition does not depend on the renderer, you are not locked to it: if you would rather not run a closed binary, you can render through Chrome instead.

It's available now on GitHub. The OpenEdit pipeline is open source under Apache-2.0. Only VEED's renderer binaries are closed source, under PolyForm Shield 1.0.0, and they are free to use: you can sell what you make with them at no cost to VEED. Transcription is built in. The default workflow gives you access to VEED's transcription when you provide a VEED account. Your audio — not the video — is uploaded and stored to transcribe it, and the limits are whatever your VEED account already has.

What OpenEdit can do

Captions are the most established path, but they are one job out of several.

Capability What that means
Edit, cut and reframe footage Trim, restructure and reframe clips for a different aspect ratio, described in a prompt rather than dragged on a timeline
Layer motion graphics and visual elements Charts on the numbers, titles, lower thirds, overlays, transitions, composed in HTML and CSS
Turn slides or websites into video Hand it a deck or a URL and get a video out. No screen recording, no manual timing
Capture web pages Pull a live page into a composition as footage
Add captions and subtitles Transcribe the audio, design a caption style that fits the footage, and burn it in
Generate its own assets Generate footage with VEED Fabric through the VEED API, or produce images, music and sound effects in whichever model suits the job. You can still point it at any other generating service or MCP server. Generating spends credits, and OpenEdit asks first

Source video is optional, and with Fabric wired in that now holds up in practice. Stills, slides, generated images or pure motion graphics are enough on their own, and your agent can fill the gaps itself rather than you sourcing assets first. Want a talking head from a script? OpenEdit can generate one with Fabric through the VEED API. Generating anything spends credits, and OpenEdit asks before it does.

Learn more about MCPs and what they do in this article.

What OpenEdit includes

Feature What it means for your workflow
No GUI, no timeline Describe the job to your agent and OpenEdit produces the video. Nothing to open, nothing to configure
Our own HTML-to-video renderer You compose in HTML and CSS, but there is no headless browser in the pipeline. Everything renders locally on your machine
Not locked to our renderer Composition does not depend on VEED’s renderer. If a closed binary is a dealbreaker, render through Chrome instead
Preview window When the render finishes, a preview window opens so you can watch the result straight away
Keep prompting Ask your agent to change what came back: darken a color, move the text, add motion graphics, match your brandbook
One-line install Run npx skills add veedstudio/open-edit then open your agent, which handles the rest of the setup itself
Open source pipeline The pipeline is Apache-2.0. Add your own APIs and MCPs, mix it into your existing skillset, connect it to any pipeline you already run
Generate what you don't have With the Fabric integration, your agent can generate footage, images, music and sound effects rather than waiting for you to supply them

How it works

OpenEdit follows a four-step flow.

Step 01: Install the skill

npx skills add veedstudio/open-edit

Step 02: Open your agent

OpenEdit is agent-agnostic. It has been tested with Claude Code, Codex and Gemini. The skill prepares the runtime and loads its AGENTS.md explicitly, so it is not tied to one model.

claude

Step 03: Describe the job

Your agent works out what to do, composes it, and renders it. Captions are one job. Reframing, motion graphics, slides, and web captures all start the same way.

Add subtitles to my video [VIDEO]

Or run the skill directly on a selected video.

/open-edit

Step 04: Review and keep going

One to three minutes later you get a captioned MP4 back, and a preview window opens. From there, keep prompting.

Make the yellow darker, move the text up a bit

Add motion graphics to the revenue numbers

Reframe this to 9x16 and keep the speaker centered

Turn this deck into a video with titles between each slide

Using the Figma MCP, take this layout and build a branded subtitle style for my podcast

Run 6 agents in parallel, each with its own caption system

What you can build with OpenEdit

A few ways OpenEdit gets used.

  • Daily social output. Ship a week of Reels, TikToks, Shorts, or podcast cuts straight from your agent, without paying for another subscription to do it.
  • Branded assets that stay on brand. Give OpenEdit a reference image or your brandbook once, and every clip after it matches.
  • A client's week, done in an afternoon. Run the batch through a multi-agent session instead of paying studio rates for the same output.
  • Video inside the tools you already run. Connect any MCP or API, make OpenEdit part of the workflow and skillset you already have, or build something new around it.

Captions are the most established path. Motion graphics and editing now sit in the skill itself, so reframing, layering and slides-to-video are jobs you can ask for directly. They are newer than captioning, so push on them. That is where the pipeline is going and where feedback is most useful right now.

How OpenEdit compares

VEED built its own HTML-to-video renderer and runs it locally on your machine. You style captions in HTML and CSS, but there is no headless browser in the pipeline. Most agent-native video tools drive Chrome instead.

In VEED's early internal benchmarks, OpenEdit ran up to 2.2x faster than tools that drive Chrome. That is VEED's own testing, not a published comparison.

Metric OpenEdit Comparable browser-rendered workflows
Render speed Up to 2.2x faster Baseline
Pipeline HTML-to-video renderer, no headless browser Headless Chrome in the render path
Where it runs Locally, on your machine Varies, often cloud or local browser

Source: VEED's own early internal benchmarks, not a published comparison. Reproducible benchmarks and the analysis behind them are coming in the next few weeks.

Good to know before you start

OpenEdit is in beta. Here is where it stands today.

  • Apple Silicon Mac on macOS 26 (Tahoe) for now. Not Intel, not earlier macOS. Windows support is arriving soon. Linux is something VEED intends to build, prioritized by demand.
  • Agent-agnostic. Tested with Claude Code, Codex, and Gemini. The skill loads its AGENTS.md explicitly, so it is not tied to one model.
  • Motion and editing now sit in the skill. Cutting, reframing and layering motion graphics are jobs you can ask for directly. Captioning is still the most established path, so expect rougher edges here. Worth pushing on regardless.
  • It now generates and edits. OpenEdit composes and renders, and with the Fabric integration it can generate footage, images, music and sound effects when a job needs them. Generating spends credits, and it asks first.

Is OpenEdit open source?

Yes. The OpenEdit pipeline is open source under Apache-2.0. Only VEED’s renderer binaries are closed source, under PolyForm Shield 1.0.0, and they are free to use: you can sell what you make with them at no cost to VEED. You are not locked to them either, because composition does not depend on the renderer, so if a closed binary is a dealbreaker, you can render through Chrome instead. Both licenses are in the repo.

What are the system requirements?

OpenEdit runs on Apple Silicon Macs on macOS 26 (Tahoe). Not Intel, and not earlier versions of macOS. Windows support is arriving soon. Linux support is something VEED intends to build, and how it gets prioritized depends on demand.

What can OpenEdit do today, and what is coming?

OpenEdit can edit, cut and reframe footage, layer motion graphics and visual elements, turn slides or websites into video, capture web pages, and add captions and subtitles. With the Fabric integration it can also generate its own assets: footage through the VEED API, plus images, music and sound effects, and it can build a soundtrack from several tracks at once. Source video is optional as a result. Generating spends credits, and OpenEdit asks first. Windows support arrives next week, along with Lip Sync v2 and background removal.

Powered by VEED Engine

What the licenses mean for you

The OpenEdit editor is open source under Apache-2.0.

The renderer binaries are closed source, under PolyForm Shield 1.0.0. You can sell what you make with them, at no cost to VEED. You are also not stuck with them: composition does not depend on VEED's renderer, so if a closed binary is a dealbreaker, you can render through Chrome instead.

Both licenses are in the repo.

Setup is one command, and your agent takes care of the rest. Open it, point it at a clip, and ask for subtitles.

Get OpenEdit on GitHub

Faq

What is OpenEdit?

OpenEdit is an open-source, agent-driven video editing pipeline that ships with VEED's HTML renderer, which is closed source but free to use. The pipeline is driven entirely through your coding agent: edit, cut and reframe footage, layer motion graphics and visual elements, turn slides or websites into video, capture web pages, and pull in any video or image generating service or MCP server when it helps. Source video is optional, since stills, slides, generated media, or pure motion graphics are enough when the brief calls for it.

How do I install OpenEdit?

Run one command in your terminal: npx skills add veedstudio/open-edit. Then open your agent and describe the job, or run /open-edit on a selected clip. Your agent handles the rest of the setup itself. If you would rather not touch the terminal, hand the command to your agent and ask it to make it work.

Which AI agents work with OpenEdit?

OpenEdit is agent-agnostic. It has been tested with Claude Code, Codex, and Gemini. The skill prepares the runtime and loads its AGENTS.md explicitly, so it is not tied to one model.

Do I need to know how to code to use OpenEdit?

No. Run one install command, then talk to your agent the way you would talk to an editor: reframe this to 9x16, add motion graphics to the revenue numbers, turn this deck into a video, and make the yellow darker. Builders who want to wire OpenEdit into a pipeline of their own can, because the pipeline is open source.

How is OpenEdit different from browser-based agent video tools?

VEED built its own HTML-to-video renderer. You style captions in HTML and CSS, but there is no headless browser in the pipeline, and it runs locally on your machine. In VEED's early internal benchmarks, it ran up to 2.2x faster than tools that drive Chrome. That is VEED's own testing rather than a published comparison. Reproducible benchmarks and the analysis are coming in the next few weeks.

What does OpenEdit cost?

OpenEdit is free to use. No subscription, no paywalls. One veed.io account for transcription. Transcription is built in. The default workflow gives you access to VEED's transcription when you provide a VEED account. Your audio — not the video — is uploaded and stored to transcribe it, and the limits are whatever your VEED account already has.

When it comes to  amazing videos, all you need is VEED

Create your first video
No credit card required