Loading…
Loading…
Agents / Short Video Agent
A voice, a picture for each scene, subtitles timed to every word and music made for it, about 17.2¢ for 30 seconds. Change any scene, then download the MP4. How it works · Prices
You can look around. To run anything, sign in (free) and add funds to your balance: you pay per use, no subscription.
Sign in to startThe script is written here, scene by scene, with a picture for each. You can change any scene when it is made.
Or paste your own script. Pick the length (15 to 90 seconds), the shape (shorts, YouTube or square), a picture style and the language.
The script split into scenes, a voice, a picture for each scene, music made to the voice's length, and the subtitles timed to every word. Usually a minute or two.
Make another picture for a scene, use your own, animate it into a moving clip, or change what a scene says. Then download the MP4, made in your browser.
“Why the ocean is salty”, 15 seconds, 9:16, photo style: the scenes it wrote.
Rainwater dissolves minerals as it flows over land, carrying ions into rivers.
Picture: Rainwater flowing over a rocky hillside
Rivers bring these salts to the sea, and evaporation removes water but leaves the salts.
Picture: A river emptying into a clear blue sea
Over millions of years, this steady flow keeps the ocean salty.
Picture: A vast ocean horizon under a bright sky
Paid per video from your wallet; no subscription. Typical prices with the default models (OpenAI TTS-1, FLUX.3 [klein], ElevenLabs Music v2).
| Length | Scenes | Typical price |
|---|---|---|
| 15 seconds | 3 | 8.7¢ |
| 30 seconds | 6 | 17.2¢ |
| 60 seconds | 12 | 34.3¢ |
| 90 seconds | 18 | 51.4¢ |
| Extras, per scene | Price |
|---|---|
| Animate a picture into an AI video clip (Hailuo 2.3 fast) | 23.9¢ |
| Another picture | 1.1¢ |
| New words (the voice is made again) |
| 0.85¢ |
| Your own picture, a subtitle style, the MP4 download | Free |
A video is charged what its parts actually cost, which depends on how long the narration is. Every part is listed on your usage page. Up to 20 videos a day.
Nothing is uploaded to TikTok, YouTube or Instagram for you. You download the video and decide where it goes.
Every picture is made for your video by an AI image model, or is your own. Music is made for it too.
Each topic and script is checked first, so it is not used for misinformation, hate or other harmful videos.
You pay for each part as it is made, from your GridPup balance: the script, the voice, the pictures and the music, at the AI models' prices. The table above shows typical prices; the exact price is shown before you start. There is no subscription.
By default a fast open model writes the script, OpenAI's voice reads it, FLUX makes the pictures and ElevenLabs makes the music. You can pick another model for each part: other voices, picture models and video models are offered, each with its price.
Yes. For each scene you can make another picture (with a new description if you like), use your own picture, animate the picture into a short AI video clip, or change what the voice says; the voice is made again so it flows. The subtitles and their style can be changed for free.
The parts are made by AI models through Cloudflare's AI Gateway and kept 30 days in your creations. The MP4 itself is put together in your browser (Chrome, Edge or Safari on a computer), so that step is free and quick.
Each video's brief and script are stored on GridPup so you can come back to them, and its files are kept 30 days in your creations. "Download my data" gives you a zip; "Delete my videos" removes them. GridPup does not train models on what you make.
Make a video. See also our privacy policy and terms.
The Short Video Agent's pipeline is based on MoneyPrinterTurbo by Harry, used under the MIT licence.