01 / 05 · AI films
IV STUDIO.
Short films with recurring AI actors. Give a character a face and a voice, hunt a few quick takes, finish the best one, then cut it all together.
Get it free
04 / Resources / Downloads
FREE TOOLS.
FROM THE STUDIO.
Five apps built for our own creative work, free for yours. No account, no ads, no card. Scroll the reel, or jump straight to one.
The shelf
PICK A TOOL.
Everything here is free and open source (MIT). Each one has its own page on GitHub.
01 / AI films
IV Studio
AI films with recurring actors, on your PC or your own cloud GPU.
Windows · Mac · Linux · NVIDIA GPU or cloud
02 / Video & audio
IV Downloader
Save video and audio from YouTube, Vimeo, X, TikTok and 1,000+ other sites.
Windows · Mac
03 / Messages
IV Chat
WhatsApp, Messenger, Instagram, LinkedIn and Telegram in one window.
Windows
04 / Dictation
IV Voice
Hold a shortcut, speak, let go: your words appear wherever you’re typing.
Windows · NVIDIA GPU or CPU
05 / Keyboard & room control
IV Keys
An on-screen keyboard that never steals focus, and a family control center.
Windows
02 / 05 · Video & audio
IV DOWNLOADER.
Paste a link, pick a format, keep it. Works with YouTube, Vimeo, X, TikTok, Instagram and more than a thousand other sites.
Get it free
03 / 05 · Messages
IV CHAT.
WhatsApp, Messenger, Instagram, LinkedIn and Telegram in one window. Your own Codex or Claude can draft replies, but nothing is ever sent for you.
Get it free
04 / 05 · Dictation
IV VOICE.
Hold a shortcut, speak, let go, and clean text appears wherever you’re typing: email, docs, prompts, code comments. Whisper runs on your own PC.
Find out more
05 / 05 · Keyboard & room control
IV KEYS.
An on-screen keyboard that never steals focus: tap keys and text lands in the app you were using. It also drives IV Voice and sets the room for Baby Mode or Work Mode.
Find out more
01 / AI films
IV STUDIO.
Short films with recurring AI actors. Give a character a face and a voice, hunt a few quick takes, finish the best one, then cut it all together.
How it works, on GitHubRecurring actors
Start with the ready-made characters Tess and Theo, or design and train your own. Put one in any shot and it’s the same face every time.
Shots, pictures, voices and music
Describe a shot and let the prompt be written for you, pictures included. Give each character a voice, type a line and they say it on screen. Make songs and scores for your edits.
From shot to film
Join shots into one continuous scene, then cut, title and export in a full-screen, Premiere-style editor. Your own agent can work alongside you over MCP.
No big graphics card, or on a Mac? Use the cloud
Shots and sequences can render on a GPU in your own RunPod or Modal account, billed by the second, with a running time and cost meter and a Shut down button, or on Comfy Cloud. On a Mac that’s how every shot renders. Pictures and music need Comfy Cloud or a ComfyUI of your own.
Your PC
- Windows 10 or 11 with an NVIDIA graphics card. It’s tested on 24 GB of video memory; 16 GB works with smaller previews.
- ComfyUI, which is free, running on the same PC.
- About 60 GB of free disk space for the video models, plus about 12 GB each for voices and music if you use them.
Install
- Download IV-Studio-Setup.exe with the button above and run it. If Windows shows “Windows protected your PC”, choose More info, then Run anyway.
- Start ComfyUI, then IV Studio. The Setup page checks everything and downloads the models it needs, each one verified before use.
- No suitable graphics card? The Cloud page sets up rendering on your own RunPod or Modal account instead, billed by the second, or on Comfy Cloud.
Mac
- Pick your Mac’s download. Not sure which? Apple menu → About This Mac. “Chip: Apple M…” means Apple Silicon.
- Open the .dmg and drag IV Studio into Applications. The first time you open it, macOS says it can’t verify the app: click Done, then go to System Settings → Privacy & Security and click Open Anyway. You only do this once.
- A Mac has no NVIDIA card for the video model, so open the Cloud page and connect your RunPod or Modal account, or Comfy Cloud. Every shot renders there.
Linux
- Download IV-Studio.AppImage, make it executable (chmod +x IV-Studio.AppImage) and run it. AppImages need FUSE: libfuse2, called libfuse2t64 on Ubuntu 24.04.
- Point it at your ComfyUI folder in Settings → ComfyUI, or let it find one in your home folder. Saving API keys needs GNOME Keyring or KWallet.
- Renders use your NVIDIA card or the cloud. DLSS 5 is Windows-only, so finals get a 2× AI upscale and frame interpolation instead.
IV Studio doesn’t include any AI models: each one downloads from its maker under its own licence. MiniMax H3’s licence doesn’t cover use in the USA, the UK, the EU or South Korea without a separate licence from MiniMax, so check it before you start. Characters and voices must be invented, your own, or of someone who agreed. The screenshots show renders made with IV Studio.
02 / Video & audio
IV DOWNLOADER.
Paste a link, pick a format, keep it. Works with YouTube, Vimeo, X, TikTok, Instagram and more than a thousand other sites.
Release notes and earlier versionsEvery format you need
Best quality up to 4K, a specific resolution, or MP3 and M4A audio. Video is saved as H.264 MP4 wherever the site offers it, so it drops straight into Premiere, Resolve or CapCut.
A queue that keeps going
Download several at once, or a whole playlist or channel. Close the app mid-download and it picks up where it left off.
Stays current
The download engine updates with one click. On Windows the app updates itself; on Mac it tells you when a new version is ready.
Windows
- Download IV-Downloader-Setup.exe with the button above.
- Run it. If Windows shows “Windows protected your PC”, choose More info, then Run anyway.
- Open IV Downloader, paste a link and pick a format.
Mac
- Pick your Mac’s download. Not sure which? Apple menu → About This Mac. “Chip: Apple M…” means Apple Silicon.
- Open the .dmg and drag IV Downloader into Applications.
- The first time you open it, macOS says it can’t verify the app. Click Done, then go to System Settings → Privacy & Security and click Open Anyway. You only do this once.
On first launch the app fetches the latest download engine, yt-dlp, from its official release and checks it before use. Download what you have the right to keep: your own work, videos whose creators allow downloads, and public-domain or openly licensed material. Respect each site’s terms. Copy-protected streaming services such as Netflix, Prime Video, Disney+ and Spotify aren’t supported. IV Downloader is built on open-source tools, including yt-dlp, FFmpeg and Deno; their licences are included in the app. The screenshot shows Blender Studio open movies, licensed CC BY.
03 / Messages
IV CHAT.
WhatsApp, Messenger, Instagram, LinkedIn and Telegram in one window. Your own Codex or Claude can draft replies, but nothing is ever sent for you.
How it works, on GitHubOne window
WhatsApp, Messenger, Instagram and LinkedIn run as their own websites inside IV Chat, each signed in separately. Telegram connects directly through Telegram’s official API, with folders, voice messages, polls and reactions.
Updates in one list
Your unread counts and the latest messages from WhatsApp, Messenger and Telegram on one page. Closing the window keeps IV Chat in the tray, with notifications.
Your AI drafts, you send
A Draft button writes a reply with the Codex app or Claude Code already on your PC, and lands it in the reply box for you to check. Connect your AI over MCP to read only the chats you allow. There’s no send tool: a message only goes when you press Enter or Send.
What you need
- Windows 10 or 11, 64-bit.
- For Telegram: your own free api_id and api_hash from my.telegram.org. IV Chat shows you where.
- For Draft reply and MCP (optional): the Codex app or Claude Code, installed and signed in.
Install
- Download IV-Chat-Setup.exe with the button above and run it. If Windows shows “Windows protected your PC”, choose More info, then Run anyway.
- Open each service in the left bar and sign in. For Telegram, add your app keys, then scan the QR code with your phone.
- Optional: open Settings → Your AI to pick which chats your AI can see.
The screenshot shows demo chats with invented people.
04 / Dictation
IV VOICE.
Hold a shortcut, speak, let go, and clean text appears wherever you’re typing: email, docs, prompts, code comments. Whisper runs on your own PC.
How it works, on GitHubTalk anywhere
Hold Ctrl + Win, speak and let go: the text is typed into whatever app you were using. Tap Space while holding for hands-free, then press the shortcut again to finish.
Private by default
Speech recognition runs locally with whisper.cpp and Whisper large-v3 turbo, so your voice never leaves the PC. PCs without a capable GPU can use the CPU build, or bring their own Groq or OpenAI key.
Spelled your way
Dictionary words are passed to Whisper as hints, so names and jargon come out right. Your history stays on your PC, and audio is never saved.
The screenshot uses sample numbers.
05 / Keyboard & room control
IV KEYS.
An on-screen keyboard that never steals focus: tap keys and text lands in the app you were using. It also drives IV Voice and sets the room for Baby Mode or Work Mode.
How it works, on GitHubType anywhere
Keys fire the moment you press them and land in the window behind, which never loses focus. Shift, Ctrl and Alt are one-shot, Caps stays on, and an emoji tray sits alongside.
Dictation built in
Start, finish, paste or cancel IV Voice dictation from big buttons.
Set the room
Baby Mode locks the physical keyboard, guards the power button and stops the Stream Deck, while the on-screen keys still type. Work Mode unlocks the keyboard, turns Filter Keys off and the Stream Deck back on.
Start a conversation
QUESTIONS OR
SOMETHING NOT WORKING?
Tell Matt which app, what you tried and what happened.