HTML-Video: How to Install and Set Up (2026 Guide)

🟢 Beginner–Intermediate   ⚙️ Type: AI Video Generator / CLI Tool   💸 Free & Open Source (Apache-2.0)   ⭐ 52 GitHub Stars


What is html-video?

html-video is an open-source tool that turns a simple text description — or even just a link to an article or GitHub repo — into a real MP4 video file, all on your own computer.

You type (or paste) what you want. An AI coding agent reads your request, picks a visual template, builds animated HTML frames, and then renders everything into a proper video using your local browser and a video encoder called ffmpeg. No monthly fees. No cloud rendering. No watermarks.

Think of it like this: you talk to an AI assistant in a browser window, it builds the video behind the scenes, and you download an MP4 when it’s done.

It comes with 21 ready-made templates covering data charts, glitch title cards, cinematic hero shots, product promos, explainer slides, and more.


Who is it for?

  • Content creators who want to turn blog posts or articles into short video clips automatically
  • Developers and coders who use AI coding agents like Claude Code, Cursor, or Codex
  • Social media managers who need fast animated video for Instagram Reels, YouTube Shorts, or LinkedIn
  • Data journalists and analysts who want to animate charts or data stories quickly
  • Open-source enthusiasts who want to explain a GitHub project as a video
  • Anyone who wants to create professional-looking videos without paying per render or learning video editing software

What makes it special?

  • Paste a link, get a video — give it a URL (article, GitHub repo, or just a prompt) and it builds the video from real content, not a generic template
  • Runs 100% on your machine — no cloud, no per-render fees, no data sent to a third party (except an optional soundtrack feature)
  • 21 license-clean templates — every template is cleared for commercial use, no legal headaches
  • Works with 6 different AI agents — Claude Code, Cursor Agent, Codex CLI, Hermes, Open Design (Vela), or just a raw Anthropic API key
  • Multi-frame storyboarding — it doesn’t just make a single slide; it builds a full multi-scene video with proper pacing
  • Optional AI soundtrack — add background music or narration using MiniMax AI, mixed directly into the MP4
  • Apache-2.0 license — fully free, no seat limits, no contributor agreements
  • Browser studio + command-line — use the visual studio in your browser, or script it from the terminal

Requirements before you start

Before installing, make sure you have these on your computer:

  • Node.js (version 18 or higher) — download here
  • pnpm — a fast package manager (different from the default npm). Install it by running: npm install -g pnpm
  • ffmpeg — a free video encoder. Download here and make sure it’s available in your terminal
  • Google Chrome or Chromium browser — the tool uses it in headless (invisible) mode to record your video frames
  • At least one AI agent OR an Anthropic API key:
    • If you have Claude Code, Cursor, or Codex installed — it will detect them automatically
    • If not, just get an Anthropic API key and paste it into the studio settings
  • Git — to clone the project from GitHub
  • 4 GB free disk space — for the project, dependencies, and rendered videos

💡 Tip: Not sure if you have these? Open your terminal (Mac: search “Terminal”, Windows: search “Command Prompt” or “PowerShell”) and type node --version, ffmpeg -version, and google-chrome --version to check.


Step-by-step installation

Step 1 — Install pnpm (if you haven’t already)

Open your terminal and run:

npm install -g pnpm

Then check it worked:

pnpm --version

You should see a version number like 9.x.x.


Step 2 — Clone the html-video repository

This downloads the project to your computer:

git clone https://github.com/nexu-io/html-video.git
cd html-video

Step 3 — Install all dependencies

This installs everything the project needs (it may take 1–2 minutes):

pnpm install

Step 4 — Build the project

pnpm -r build

Wait for this to finish. You’ll see build output scrolling by — that’s normal.


Step 5 — Run the doctor check

This checks that all your tools (ffmpeg, Chrome, AI agents) are detected correctly:

node packages/cli/dist/bin.js doctor

You’ll see a list of what’s found and what’s missing. Fix any missing tools before moving on.


Step 6 — Start the studio

node packages/cli/dist/bin.js studio

This starts a local web server. You’ll see a message like:

Studio running at http://127.0.0.1:3071

Open that link in your browser (Chrome is recommended).


Step 7 — Set up your AI agent or API key

In the studio, look at the top bar for the agent switcher. If you have Claude Code, Cursor, or Codex installed, it will already appear there. If not:

  1. Click the Settings icon
  2. Find the Anthropic API Key field
  3. Paste your key (starts with sk-ant-...)
  4. Save — you’re ready to go

Step 8 — Create your first video

  1. In the studio, browse the template gallery and click one you like
  2. In the chat box, describe what you want — for example: “Make a 30-second explainer video about how solar panels work”
  3. Or paste a URL: “Turn this article into a video: https://example.com/article”
  4. The AI agent will build the storyboard, fill in the content, and show you a preview
  5. Edit any frame text you want to change
  6. Click Export MP4 — your video renders locally and downloads automatically

Optional: Add AI music or narration

If you want background music or a voiceover mixed into your video:

  1. Get a free MiniMax API key
  2. Go to Settings → Audio in the studio
  3. Paste your MiniMax key
  4. In your project’s Soundtrack panel, describe the music mood or type a narration script
  5. Export — music and voice will be mixed into the final MP4 automatically

Common errors and fixes

ErrorWhat it meansHow to fix it
pnpm: command not foundpnpm is not installedRun npm install -g pnpm first
ffmpeg: command not foundffmpeg is missing or not in your PATHDownload ffmpeg from ffmpeg.org and add it to your system PATH
Chrome not foundThe tool can’t find Chrome or ChromiumInstall Google Chrome, or install Chromium. On Mac you may need to run: brew install --cask chromium
No agent detectedNo AI coding agent found and no API key setAdd your Anthropic API key in Settings → the studio will use the API directly
Build failed / TypeScript errorNode.js version is too oldRun node --version. If it shows v16 or lower, upgrade to Node.js 18 or higher from nodejs.org
Port 3071 already in useSomething else is already running on that portClose other local servers, or restart your computer and try again
Video renders but has no soundMiniMax key not configuredAdd your MiniMax API key in Settings → Audio. Audio is optional — the rest works without it
Article URL doesn’t load contentThe page may require a loginThe tool can only fetch publicly accessible pages. Try copying the article text into the chat box instead

Free vs Paid comparison

Featurehtml-video (Free)Paid Video AI Tools
Cost per video$0 (renders locally)$5–$50+ per month
Monthly render limitsUnlimitedOften capped (e.g. 20–100/month)
WatermarkNoneOften on free tiers
Your data stays private✅ Yes — runs on your machine❌ Sent to cloud servers
Commercial use✅ Apache-2.0 — fully allowedVaries — check licence carefully
Requires technical setup⚠️ Yes — Node.js, ffmpeg, Chrome❌ No — browser only
Customisable templates✅ Yes — full HTML/CSS controlLimited to built-in options
AI agent integration✅ 6 agents supportedUsually none
AI soundtrack✅ Via MiniMax API (BYOK)Sometimes included
Ease of use🟡 Moderate — needs CLI setup🟢 Very easy — no setup

Bottom line: If you’re comfortable with a terminal and want unlimited free renders with full privacy, html-video wins. If you want a no-setup browser tool and don’t mind paying, a cloud service may suit you better.


Alternatives — 3 similar tools

1. Remotion

A React-based video framework for developers. You write your video in code (React components) and it renders to MP4. Very powerful and polished, but has a paid licence for teams above 4 developers. Great if you already know React.

🔗 www.remotion.dev

2. Motion Canvas

An open-source, code-first animation library perfect for explainer videos, code walkthroughs, and math visualisations. You write animations in TypeScript. Completely free. Best for developers who want precise control over every frame.

🔗 motioncanvas.io

3. Manim (by 3Blue1Brown)

The engine behind the famous YouTube channel 3Blue1Brown. Specialises in mathematical and scientific animations. Completely free and open-source, but has a steep learning curve and is best suited to Python developers who need clean data or math visualisations.

🔗 www.manim.community


🚀 Want more free AI tools like this?

We find, test, and write setup guides for the best free and open-source AI tools — so you don’t have to dig through GitHub yourself.Browse Free AI Tools at globalaiforce.com/browse-tools/→


📸 Follow us for daily AI tool tips and tutorials: instagram.com/globalaiforce

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top