01
Type
Write or paste your reply in the floating panel, up to 280 characters. Try typing here. The video below is a real SayWords render.
Reply length0 / 280
Reply0/280
Type a reply. Get a video of you saying it, in your own voice without internet.
Not affiliated with or endorsed by X Corp., Reddit, Meta, Slack or Discord. Trademarks belong to their owners.
Text replies get lost in the timeline. Recording a camera video takes setup, retakes and editing. SayWords sits in between: a reply with a face and a voice, made in about the time it takes to type it.
Fully offline. No internet, no cloud. Your face, voice and videos never leave your Mac.
No export button. Skip the rendering time and drag the video straight into your reply box.
Upload your face once. It's rigged into every animation automatically.
50+ animations. Coffee, walking, cycling, biking and more. A move for every mood.
Customize almost everything. Colours, backgrounds and caption styles, all your way.
Your own cloned voice. It sounds like you recorded it and hit send yourself.
Pay once, own it for life. No subscription.
Each word is highlighted as your voice speaks it, so the reply still works when it plays with the sound off.
Upload one photo. SayWords cuts out your head and places it on a hand-drawn character that moves through 50+ animations.
Give it a short sample once. A local Python TTS engine then speaks whatever you type in your voice.
H.264 MP4 in two wide shapes, made to sit inline in a reply thread.
Exported at the exact ratio with no padding, so it plays inline in the feed on X, Reddit and Threads.
Clean pastel palettes, Grid, Dots, Solid or Gradient patterns, and Geist, Bricolage or Space Mono type.
A native floating launcher that stays above your other windows, right next to your browser.
Built on macOS drag and drop (NSItemProvider). Drag the file straight out of the window, with no save dialog and no file picker.
Plain MP4 output for X, Reddit, Threads, Slack and Discord. SayWords is independent and not affiliated with or endorsed by X Corp., Reddit, Meta, Slack or Discord.
Write or paste your reply in the floating panel, up to 280 characters. Try typing here. The video below is a real SayWords render.
Pick one of 50+ animations that fits the tone. Voice, captions and render happen in the background while you type.
Drop the MP4 into the composer. The reply text is already copied to your clipboard.
Product specs, not usage stats. We don't have public user numbers yet.
Type the reply, pick a pose and drag the MP4 into the reply box on X, Reddit or Threads. The text goes to your clipboard so you can paste it as the caption.
Read the guideIt's a regular MP4, so you can drop it into a Slack thread, a Discord channel or any app that accepts a video file. SayWords isn't affiliated with any of these platforms.
Video messages for SlackTwo lifetime plans, paid once. No subscription and no credits. Pick $4.99 with a small watermark, or $13 for clean videos.
Works on macOS 13 Ventura or later, on Apple silicon and Intel Macs.
Secure checkout by Dodo Payments.
A lifetime licence to SayWords for macOS, paid once. Videos on the $4.99 plan carry a small "Made with SayWords" watermark. The $13 plan removes it.
Both are one-time lifetime licences with the same features. The only difference is the watermark: $4.99 adds a small "Made with SayWords" mark to each video, and $13 gives you clean videos.
More than 50, including sipping coffee, walking, cycling, riding a bike and a heart gesture. Your face is placed on every one automatically.
It's built for replies on X, Reddit, Threads, Slack and Discord. The output is a regular MP4, so you can drag it anywhere that accepts a video file, including Chrome, Safari and Finder. SayWords isn't affiliated with any of these platforms.
SayWords exports at exactly 3:1 or 2.5:1, with no padding added. In the X feed it plays inline with no black bars, on desktop and mobile. If someone taps it open full-screen on a phone, the player centres it, which is normal for any wide video.
You give it a short audio sample once, and a local Python text-to-speech engine speaks your typed reply in your voice. Only clone your own voice.
They stay on your Mac. Your face, voice sample, typed text and videos are processed offline and never sent to a server. The internet is only used once, to activate your licence. You can remove a face or voice profile inside the app, or delete the ~/SayWords folder in Finder.
For a typical 280-character reply, here's roughly how long it takes until the MP4 is ready to drag or paste:
| Mode | Apple silicon (M1–M4) | Intel (Core i5 / i7) |
|---|---|---|
| Without voiceSilent animated video | 0.8–1.2 s | 1.5–2.5 s |
| With voiceYour cloned voice + video | 2.5–3.5 s | 7–10 s |
Without voice, it just draws the frames and encodes the video, so it's practically instant. With voice, your voice is generated on your Mac first, then synced word by word with the video. No cloud queue either way. Times are approximate and depend on your Mac.
Any Mac on macOS 13 Ventura or later, including Sonoma and Sequoia. It runs on both Apple silicon (M1 to M4), where it uses the GPU for faster rendering, and Intel Macs.
Type it like a comment. Type it. Say it. Ship it.
Claim my lifetime licence from $4.99 $4.99 with watermark · $13 without