P-Video is Pruna's video generation model — text-to-video, image-to-video, and audio-to-video modes. Optimized for cost efficiency and fast turnaround. Strong on cinematic landscapes, character beats, and product macros.
P-Video is Pruna's video generation model — designed for cost-efficient turnaround while preserving production-grade quality. Three modes: text-to-video (compose from prompt), image-to-video (animate a starting frame), audio-to-video (drive motion from an audio track).
Two resolution tiers (720p, 1080p), six aspect ratios covering cinematic through vertical, 24 / 48 FPS options, and an optional draft mode for ultra-cheap iteration. Strong on cinematic landscapes, character beats, product macros, and atmospheric subjects.
3
Modes — T2V · I2V · A2V
6
Aspect ratios
24 / 48
FPS options
P-Video supports three modes. Text-to-video composes from prompt alone. Image-to-video animates a starting frame with motion prompts. Audio-to-video drives motion from an audio track (lip-sync, music-driven motion, dance choreography).
720p is the production sweet spot — strong motion fidelity at low credit cost. 1080p for hero shots and final renders where every detail counts. Pruna's optimization keeps both tiers cost-efficient relative to other video models.
24 FPS is the cinematic standard — what feature films use. 48 FPS doubles the frame rate, useful for slow-motion downstream edits (50% playback speed produces 24 FPS slow-mo) or smoother motion in fast action.
Draft mode produces lower-fidelity output at significantly lower credit cost — perfect for rapid prompt iteration. Get the composition right at draft, then re-render at production fidelity. Don't ship draft mode; iterate with it.
Each example shows the exact prompt that produced the result. Copy any prompt with one click.
720p · 16:9 · 5s · 24 FPS
A panoramic drone shot moves slowly across a misty Scottish Highland valley at sunrise, golden light catching the bracken, a single hiker walks along the ridgeline with a backpack, painterly atmospheric
"Panoramic drone shot moves slowly" specifies one camera move. P-Video holds smooth drone motion cleanly. The single hiker + golden light gives a focal element to anchor the vast valley. 24 FPS reads as cinematic film standard.
720p · 16:9 · 5s · 24 FPS
A young man in a tweed jacket sits on a sunlit Brooklyn rooftop reading a hardcover novel, gentle breeze ruffles the pages, the camera slowly pushes in toward him, soft natural lighting, ambient city sounds
"Camera slowly pushes in toward him" is a classic emotional-beat move. "Gentle breeze ruffles the pages" gives a micro-motion to anchor the scene beyond the camera push. Specific outerwear ("tweed jacket") gives character identity in 5 seconds.
720p · 1:1 · 5s · 24 FPS
Macro slow-motion of a single vintage pocket watch being placed on dark velvet, the second hand begins to tick, soft warm spotlight catching the engraved case, premium timepiece commercial mood
1:1 macro product spot. "Second hand begins to tick" gives one specific motion that elevates the clip past static. P-Video renders engraved metal and velvet drape cleanly. "Premium timepiece commercial" anchors aesthetic without over-specifying.
720p · 9:16 · 5s · 24 FPS
A surfer paddles out through gentle morning waves toward an offshore lineup at sunrise, vertical handheld shot from waist deep, golden alpenglow on the horizon, atmospheric ocean ambient
9:16 vertical for mobile-social. "Vertical handheld shot from waist deep" specifies aspect AND perspective AND distance. Atmospheric ocean ambient cue helps the native audio pass — P-Video layers convincing wave/ocean sound.
720p at 24 FPS is the cinematic-standard production sweet spot. Strong fidelity, reasonable cost, reads as filmic. Bump to 1080p for hero shots; use 48 FPS only when you specifically need smoother motion or slow-motion downstream.
Draft mode produces lower-fidelity output at significantly lower credit cost. Use it for prompt iteration (5-10 attempts at low cost) until you've nailed the composition. Then re-render without draft mode for the final keeper.
"Panoramic drone shot moves slowly", "camera slowly pushes in toward him", "vertical handheld shot from waist deep" — P-Video reads camera language literally and executes one move cleanly. Combining motion types usually produces compromised motion.
"Gentle breeze ruffles the pages", "second hand begins to tick", "golden alpenglow shifts on the horizon" — add ONE micro-motion element beyond the camera move. Elevates clips past static beats with minimal prompt complexity.
Pass an image URL for image-to-video mode to anchor specific character identity, composition, or palette while your prompt describes motion. Useful for animating Pruna P-Image outputs, character work, or specific compositions.
P-Video's audio-to-video mode generates motion synced to an input audio track. Useful for music-driven visuals, dance choreography, or lip-sync work. Pair with a clean audio source for best sync results.
| Setting | Values | Notes |
|---|---|---|
| Mode | Text-to-video · Image-to-video · Audio-to-video | I2V requires a starting image. A2V requires an audio source. T2V composes from text alone. |
| Resolution | 720p · 1080p | 720p is the production sweet spot. 1080p for hero shots. |
| Aspect ratio | 16:9 · 9:16 · 1:1 · 4:3 · 3:2 · 2:3 | Six aspect presets covering cinematic, vertical, square, traditional. |
| FPS | 24 · 48 | 24 for cinematic standard. 48 for slow-motion or smoother fast action. |
| Duration | Variable seconds | Credit cost scales with duration. |
| Draft | Boolean (optional) | Lower fidelity at much lower cost. For iteration only. |
| Pricing | Pruna cost-efficient | Lower per-second cost than many premium video models. |
P-Video is in Pruna's family of cost-efficient models — designed for production-grade quality at lower credit cost than premium video models. Kling 2.6 Pro, Runway Gen-4.5, and Veo 3 produce more cinematic output at higher cost. WAN 2.7 and LTX 2.3 have broader feature surface. Use P-Video when cost-per-second is the binding constraint.