P-Video: Text & Image to Video Online
Bottom line
Pruna AI's efficiency-focused generator: 4-10 second clips at 720p or 1080p, from roughly 1.6 credits/second. Start from a prompt or a still image in the Text to Video tool; VdoBloom shows the exact credit cost before each P-Video run, with no separate P-Video account.
P-Video comes from Pruna AI, a team whose specialty is making AI models run leaner and faster -- and that focus is the whole point of this one. The catalog bills it as real-time text and image to video, and in practice it behaves like a sketching tool: prompts and stills turn into 4 to 10 second clips at a price low enough that repeated attempts are cheap.
It covers both directions in a single place: type a prompt for text-to-video, or attach an image to animate -- same model, same settings surface. Output lands at 720p or 1080p across seven aspect ratios (16:9, 9:16, 1:1, 4:3, 3:4, 3:2, 2:3), so drafts made here can match the framing of whatever finishing model you hand them to. Pricing runs roughly 1.6 credits/second at 720p and 3.2 credits/second at 1080p -- a 5-second 720p clip is 8 credits, well under most other video models on VdoBloom.
At a 4.5 rating it isn't the directory's top scorer, and it carries a HOT content label, VdoBloom's most permissive filtering tier. Credit cost appears before every generation, and a failed render is refunded automatically, which combined with the low per-second price makes it a practical model for volume iteration before moving a concept to a premium model.
What it costs on VdoBloom
P-Video is priced per second: roughly 1.6 credits/second at 720p, 3.2 credits/second at 1080p. Worked examples: 4s/720p = 7 credits, 5s/720p = 8 credits, 10s/720p = 16 credits; 4s/1080p = 13 credits, 5s/1080p = 16 credits, 10s/1080p = 32 credits. That's among the lowest per-second rates of any video model on VdoBloom.
Where P-Video fits next to the Wan family
P-Video and Wan aren't the same model line -- P-Video comes from Pruna AI, a team focused on inference efficiency, while Wan is Alibaba's line. On VdoBloom the practical comparison is cost and range: P-Video is cheaper per second (roughly 1.6-3.2 credits/second) but caps at 10 seconds and 1080p, while Wan 2.6/2.7/3.0 add longer durations, native audio on some variants, and higher resolution ceilings at a higher per-second price. P-Video is the model to reach for when you're iterating on a concept fast and cheap; the Wan tiers are the step up once you've settled on a direction.
What creators use P-Video for
- Rapid previsualization -- blocking out shot ideas from text prompts before committing credits to slower premium models.
- Testing an image-animated version and a text-only version of the same concept side by side without switching models.
- Generating cheap draft variants of a concept across all seven supported aspect ratios before picking one to finish.
- Producing short 4-10 second clips for formats where a longer, pricier render isn't necessary.
How to use P-Video on VdoBloom
- 1Open the Text to Video tool and select P-Video from the model picker.
- 2Write your prompt — or upload a starting image for image-to-video.
- 3Choose your duration (4–10s), quality (720p/1080p), aspect ratio, review the credit cost, and generate.
Frequently asked questions
How much does P-Video cost on VdoBloom?
Roughly 1.6 credits/second at 720p, 3.2 credits/second at 1080p. A 5-second 720p clip is 8 credits, 10 seconds is 16; at 1080p those are 16 and 32 credits.
Do my 10 free signup credits cover a P-Video render?
Yes, comfortably -- a 5-second 720p clip is only 8 credits, leaving 2 credits from a new account's free balance. It's one of the cheapest models on VdoBloom to test with free credits.
Does P-Video handle both text-to-video and image-to-video?
Yes -- it's a dual-capability model. You can generate from a prompt alone or attach a still image to animate, with the same duration, resolution and aspect-ratio controls in both modes.
What clip lengths and output sizes does P-Video support?
4 to 10 seconds in whole-second steps, rendered at 720p or 1080p, across seven aspect ratios including 16:9 widescreen, 9:16 vertical and 1:1 square.
How fast is P-Video?
The catalog lists it as real-time text and image to video, reflecting Pruna AI's specialty in model efficiency. VdoBloom doesn't publish exact render times, so treat that as directional rather than a guaranteed speed.
What content rules apply when using P-Video?
It sits in VdoBloom's HOT tier, meaning its filtering is more flexible than standard. Platform terms still govern what you can generate, but the model is less likely to refuse borderline creative prompts than STRICT-tier alternatives.
How does P-Video compare to Wan 2.2 A14B for short clips?
P-Video is meaningfully cheaper for equivalent short clips: a 5-second 720p render is 8 credits on P-Video versus a flat 36 credits on Wan 2.2 A14B (which doesn't get cheaper for shorter durations). Wan 2.2 A14B goes as short as 1 second; P-Video's floor is 4 seconds. For quick, low-cost iteration in the 4-10 second range, P-Video is the more economical pick.
Is P-Video open source?
There's no public information establishing P-Video's weights as open source, so this isn't something VdoBloom can confirm either way. It's accessed on VdoBloom as a hosted model regardless.