Pruna P Video Upscale: raise any clip to 720p, 1080p, 2K or 4K
Pruna P Video Upscale takes a clip and returns a higher-resolution H.264 MP4 — 720p, 1080p, 2K or 4K — keeping its aspect ratio, audio and frame rate, or interpolating to a smoother one.
Pruna P Video Upscale is an AI video upscaler from Pruna AI. Give it a clip and a target — 720p, 1080p, 2K or 4K — and it returns a sharper, higher-resolution H.264 MP4 without changing the content. It keeps the source aspect ratio, copies the audio, and keeps the frame rate unless you raise it. Use it to finish AI-generated clips, upgrade archival footage, or clean up compressed exports.
ModelPruna P Video UpscaleUpscale any video to 720p, 1080p, 2K or 4K.What it does well
Clean upscaling to your target resolution
Pick a target and the model enlarges the frame and sharpens fine detail, leaving the scene itself unchanged. In testing, a 720p potter's-wheel clip came back crisp at 1920×1080: the wet clay ridges and water sheen read clean, the wheel kept turning, and the studio audio stayed in sync. The same pass runs up to 4K when you need finishing resolution, and a clip already larger than the target is returned as is rather than shrunk.
Any aspect ratio, kept
Targets are pixel-area presets, so a clip that isn't 16:9 keeps its own shape — no stretching, letterboxing or cropping. A vertical 9:16 skatepark clip upscaled to 1080p came back upright at 1080×1920 with the carve still reading clearly. Phone and social footage keeps its frame automatically, so the upscaled clip drops straight back into a vertical edit.
Examples to try
Here is one more run from testing to try on your own footage.
The same clip sharpens to a full 3840×2160 frame with its 16:9 shape and audio kept. 4K processes about 10 seconds at 30 fps, so trim longer clips or pick a lower target when a job is rejected.
Recommended settings
| Setting | Value | When to use it |
|---|---|---|
| target | 1080p (default) | Everyday web and social delivery. |
| target | 2K or 4K | Large screens and finishing work; allows shorter clips. |
| fps | empty (default) | Keep the video's own frame rate. |
| fps | 48–60 | Interpolate smoother motion on steady or moderate footage. |
| turbo | off (default) | Cleanest result. |
| turbo | on | A faster draft pass with a slight quality trade-off. |
Good to know
- An upscaler sharpens detail that exists but cannot invent what was never captured, so heavily compressed or very blurry sources gain less.
- Each target has a processed-frame cap (720p 2400, 1080p 1200, 2K 600, 4K 300 frames). A clip too long for the chosen target is rejected with a message naming the limit — use a shorter clip, a lower target, or a lower frame rate.
- Raising fps interpolates frames, which can show artifacts on very fast motion.
- It is a faithful upscaler, not a restyler or generative editor; it raises resolution without changing the scene.
- Input is one video URL (mp4, mov or webm) up to 512 MiB; the output is a standard H.264 MP4.
Use it from the API
Send a public video URL and a target; the block below renders the Python and cURL to call the model. Leave fps empty to keep the source cadence, or set it up to 60 to interpolate smoother motion.
# pip install "segmind>=1.1.0"
# export SEGMIND_API_KEY="YOUR_API_KEY"
import segmind
result = segmind.run(
"p-video-upscale",
video="https://segmind-resources.s3.amazonaws.com/input/p-video-upscale-source-main-1793251047.mp4",
target="1080p",
turbo=False,
)
print(result.get("output"))# 1. Submit the request; the reply carries its request_id
REQUEST_ID=$(curl -s -X POST "https://api.segmind.com/v2/p-video-upscale" \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"video": "https://segmind-resources.s3.amazonaws.com/input/p-video-upscale-source-main-1793251047.mp4",
"target": "1080p",
"turbo": false
}' | jq -r .request_id)
# 2. Check its status until it is COMPLETED
curl -s "https://api.segmind.com/v2/requests/$REQUEST_ID/status" \
-H "x-api-key: YOUR_API_KEY"
# 3. Fetch the result
curl -s "https://api.segmind.com/v2/requests/$REQUEST_ID" \
-H "x-api-key: YOUR_API_KEY"Resources
- Try Pruna P Video Upscale on Segmind
- API reference
- p-video-upscale - Pruna Developer Portal — Pruna's official docs: quickstart, the full parameter set, per-target frame limits and how to call the model.
- Pruna API Reference — Pruna's predictions and files endpoints and the authentication header.
- #video-to-video
- #pruna-ai
- #model-launch
- #video-upscaling
- #4k
