Video GenerationActive
pruna / p-video-2-pro
p-video-2-pro
Generate 5–15 second videos from a prompt or first and last frame images. Choose Speed or Quality, and export 480p or 768p video at 24 fps with synced audio.
Text to VideoImage to Video
Model ID
p-video-2-pro
Provider
pruna
Updated
1789653609
wiro playground—pruna/p-video-2-pro
Updated 1789653609
Overview
P-Video 2 Pro is a video generation model from Pruna. It turns a written scene description into a short clip, and it can also animate from a provided first frame. You can anchor the ending with a last-frame image for tighter continuity. The model renders video and synchronized audio together, so you can prompt dialogue, music, and sound effects in one pass.
What you can build
- Cinematic B-roll clips for ads, trailers, and product pages
- Character scenes with spoken dialogue and matching lip motion
- Social shorts in vertical or square formats
- Motion shots from a single still image, like “bring this photo to life”
- Bracketed transitions by supplying both a starting frame and an ending frame
- Mood pieces with ambient sound, music beds, and effects prompted in the same text
Inputs
- A required scene description in plain text, up to 2048 characters. Include subject, action, setting, camera movement, lighting, and audio cues.
- A required clip length selection of 5, 10, or 15 seconds.
- A required output resolution choice of 480p or 768p.
- A required generation recipe choice: Speed for faster drafts, or Quality for stronger coherence and detail.
- An aspect ratio selection for text-only generation, including 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, and 1:1.
- An optional first-frame reference image (JPG or PNG). When you provide it, the clip uses that image’s canvas shape.
- An optional last-frame reference image (JPG or PNG) to steer the final composition.
- An optional prompt expansion setting. It can leave your text unchanged, expand it moderately, or expand it aggressively.
- An optional integer seed from 0 to 2,147,483,647 to repeat a result more reliably.
Outputs
- One generated video file in MP4 format.
- The clip includes an audio track generated to match the prompt.
- Video plays back at 24 fps.
- Frame size follows the selected 480p or 768p tier.
- If you provide a first-frame image, the output framing follows that image’s aspect ratio.
Recommended settings
- Pick Quality mode for anything you plan to ship. Use Speed mode for exploration.
- Keep prompt expansion on the middle setting for most work. It tends to hold the subject better than turning it off.
- Avoid the strongest prompt expansion when you need strict fidelity. It can introduce details you did not request.
- Use a seed when you need a repeatable iteration loop. Keep the seed fixed while you adjust one variable at a time.
Limitations
- Output duration is limited to 5, 10, or 15 seconds.
- Output resolution is limited to 480p or 768p.
- Output frame rate is fixed at 24 fps.
- When you provide a first-frame image, the aspect ratio selector does not control the canvas.
- First and last frame conditioning works best when the two images match the same subject and style.
- Low-quality, blurry, or heavily compressed reference images can cause drifting detail and texture breakup.
- Busy scenes with many small elements can reduce stability across frames.
- Prompt expansion can change wording and add assumptions. Review results when strict prompts matter.
Safety & compliance
- Only upload images you own or have rights to use.
- Don’t generate or distribute content that violates privacy or consent.
- Avoid impersonation and deceptive deepfakes of real people without explicit permission.
- Follow applicable laws and the platform’s content rules for generated video and audio.
Example prompts
Great starting points for p-video-2-pro.
The wind rises and the scarlet gown billows and ripples across the frame, her hair lifts and settles, she turns her head slowly toward the sea and takes one small step forward, camera pushes in gently while drifting slightly upward, golden backlight glowing through the silk, waves rolling and breaking far below, smooth continuous motion, no morphing, ambient wind and soft distant surfVideo Generation
Milk pours in a steady white ribbon into the espresso and a rosetta pattern forms across the crema, steam curling upward, the pitcher lifting away at the end, camera easing closer to the cup rim while orbiting a few degrees to the right, window light shifting across the porcelain, realistic liquid motion, ambient cafe murmur and the faint hiss of a steam wandVideo Generation
He finishes the air, grabs the board rail, rotates and drops back into the bowl, wheels touching down with a sharp clack and rolling away across the concrete, dust kicking up behind him, camera tracks him down the transition and tilts to follow, late sun flaring briefly across the lens, handheld documentary energy, ambient wheel rumble echoing in the bowlVideo Generation
The waterfall thunders down in continuous sheets, mist blows across the ledge and beads on her jacket as she lifts her hood and looks up toward the crest, camera rises slowly and pushes in slightly, spray catching the diffuse light, moss trembling in the wind, realistic continuous motion, ambient roar of falling waterVideo Generation
API quick start
Run p-video-2-pro with a single API call.
POST https://api.wiro.ai/v1/Run/pruna/p-video-2-pro
{
"prompt": "The wind rises and the scarlet gown billo…",
"duration": 5,
"resolution": "768p",
"mode": "quality"
}curl
curl -X POST "https://api.wiro.ai/v1/Run/pruna/p-video-2-pro" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"prompt": "The wind rises and the scarlet gown billo…",
"duration": 5,
"resolution": "768p",
"mode": "quality"
}
JSON