P-Video-2

P-Video-2 is Pruna’s premium video generation model, the quality-focused successor to P-Video. Generate from text, an image, or audio-conditioned input in one endpoint. Compared with P-Video, the quality jump is native-speech lip-sync (pause / line / rest), sharper close-ups, and identity lock.

Designed for production results, it offers text-to-video, image-to-video, and audio-conditioned generation at up to 1080p and 48 fps, with optional duration up to 20 seconds (or leave duration empty and the model chooses length from the prompt). Use draft mode for faster, lower-cost iteration, then switch to full quality for finals.

Note

When using P-Video-2, make sure to respect the copyright of the images and audio you use as input and the content you generate.

Pricing: billed per second of the video returned (measured from the finished clip in whole seconds).

Resolution	Draft mode is OFF	Draft mode is ON
720p	$0.025 per second of output video	$0.015 per second of output video
1080p	$0.05 per second of output video	$0.03 per second of output video

Speed: model execution time per second of output video.

Resolution	Draft mode is OFF	Draft mode is ON
720p	~0.91 s per second of output video	~0.41 s per second of output video

Tip

Test it now in the P-Video-2 Playground.

Prompt formula

Fast pass

One prompt: subject, action, scene. Enough for first looks. Iterate in draft.

[prompt]

subject, action, scene

Locked-in

Add camera, lighting, style, and audio. Optional still or imported track for repeatable runs.

[prompt] [image] [audio]

subject, action, scene + camera, lighting, style, dialogue, optional still for I2V, native speech or imported track

Fast pass: text-to-video; subject, action, and scene only.

Locked-in: text-to-video; camera, lighting, style, and audio spelled out.

prompt: required. Name the subject, action, and scene. Add camera, lighting, style, and audio when you need a repeatable final.
- Fast pass: “Knitted purple prune character, reading a book, cozy room.”
- Locked-in: “Knitted purple prune character, reading a book, eyes look at book then at camera, cozy room, camera slowly zooming in, warm natural light, documentary style. Character reading aloud: ‘P-Video-2 generates premium AI video with sharp lip-sync and native audio.’”
image: optional still for image-to-video (jpg, jpeg, png, webp). Keep motion subtle and aligned with the reference frame. aspect_ratio is ignored when an image is set. Optional last_frame_image for end-frame control.
- Fast pass: omit for text-to-video.
- Locked-in: a high-quality, well-lit still; generate it with P-Image-Ideogram when you need a locked composition.
audio: optional imported track (flac, mp3, wav). Duration follows the audio and duration is ignored. Lead lip-sync demos with native speech (save_audio: true, no imported audio).
- Fast pass: save_audio: true; name dialogue or music in the prompt.
- Locked-in: paste full lyrics when singing; import audio only when you need a specific track.

Slot	Fast pass (enough to run)	Locked-in (stronger control)
`prompt`	Subject, action, scene.	Subject + action + scene + camera, lighting, style, audio; same wording every rerun.
`image`	Empty (text-to-video).	Well-lit still; optional `last_frame_image`.
`audio`	Native speech in the prompt; `save_audio: true`.	Imported track when timing must match; otherwise keep native speech for lip-sync.

Tip

For comprehensive video prompting (motion, framing, atmosphere), see the Video Generation guide.

Choosing the right video model

Pruna ships performance video models that share the same prediction API, but each solves a different production problem. P-Video-2 generates new footage (Model: p-video-2).

	P-Video-2-Pro	P-Video-2	P-Video	P-Video-Avatar	P-Video-Animate	P-Video-Replace	P-Video-Edit
One-line job	Generate cinematic footage from prompts	Generate premium footage from prompts	Generate fast / affordable footage	Speak from one still (script or audio)	Retarget one still with clip motion	Swap characters in existing footage	Rewrite content in existing footage
You start with	Text prompt (+ optional first / last frame)	Text prompt (+ optional image / audio)	Text prompt (+ optional image / audio)	Portrait still + `voice_script` or `audio`	Source video + one still	Source video + identity stills	Source video + text prompt
You keep from the source	N/A (new scene); canvas follows reference images when set	N/A (new scene); strong I2V consistency when an image is set	N/A (new scene)	Aspect ratio of the still	Motion, timing, camera from the driver	Camera, timing, blocking, background	Camera, timing, subject performance
Typical ask	“Make an 8 s cinematic product or documentary clip with generated audio.”	“Make a 10 s product ad with music and lip-sync.”	“Draft a 5 s social clip cheaply and fast.”	“This spokesperson says this line in French.”	“Animate this catalog still using our winning ad take.”	“Put our creator in this UGC b-roll.”	“Turn this silver SUV red and keep the camera move.”

Quick decision guide

Need the highest generation quality, lip-sync, or audio-conditioned output → P-Video-2.
Need the fastest / cheapest generation iteration → P-Video.
Single-speaker talking head, voice only, no music bed → P-Video-Avatar.
Footage exists and the hero still should move like the driver → P-Video-Animate.
Footage exists and you need different people in the same shot → P-Video-Replace.
Footage exists and you need a color, product, environment, object, or text edit → P-Video-Edit.

Key features

Marketing and ads: Text-to-video social ads, image-to-video from brand stills, and clips with music or SFX. Prefer P-Video-Avatar only for voice-only talking heads.
Media and entertainment: Music visuals, short-form entertainment, and audio-driven performance clips. Paste full lyrics when singing; duration follows imported audio.
Retail and e-commerce: Product loops and shoppable lifestyle clips. Keep motion subtle and start from a clean product or on-model still.
Corporate and education: Training, explainers, and comms that need a music bed, ambience, or two speakers. Voice-only single-presenter scripts → P-Video-Avatar.
Gaming: Trailers, world reveals, and character moments. One hero action per clip; prefer 24 fps at 1080p.
Native-speech lip-sync: When the model generates the voice (save_audio: true, no imported audio), it pauses, delivers the line, then rests, and the mouth follows.
Image-to-video + last frame: Animate a reference still; optional last_frame_image for end-frame control.
Draft mode: Faster, lower-quality preview at half the per-second rate. Iterate with draft: true, then finalize with draft: false.

Practical constraints

Maximum duration is 20 seconds.
When audio is provided, output length and billing follow the audio; duration is ignored.
aspect_ratio is ignored when an input image is provided.
Input images: jpg, jpeg, png, webp. Input audio: flac, mp3, wav.
Draft mode is a preview: faster and cheaper, lower quality.
Above two speakers, speaker separation can degrade.

Examples

Source image UGC skincare source still

Output

UGC skincare hook · image-to-video · 9:16 · 1080p · 6s · voice + music bed

Example prompt

The woman holds the serum bottle toward camera and speaks enthusiastically: "Okay this serum actually fixed my dry patches in one week." Subtle head movement, natural blinking, static phone-camera framing, warm bathroom background unchanged. Audio: soft trending lo-fi hip-hop beat playing under her voice, her voice clear and close-mic, very light bathroom room tone — never voice-only.

The woman holds the serum bottle toward camera and speaks enthusiastically: "Okay this serum actually fixed my dry patches in one week." Subtle head movement, natural blinking, static phone-camera framing, warm bathroom background unchanged. Audio: soft trending lo-fi hip-hop beat playing under her voice, her voice clear and close-mic, very light bathroom room tone — never voice-only.

Output

Mascot reaction ad · text-to-video · 9:16 · 720p · 5s · music + SFX

Example prompt

Cartoon green plush bird with a chef hat, round body with no legs, sitting on a wooden kitchen table next to a neat stack of golden pancakes with syrup. Static camera, eye-level, shallow depth of field. The bird bobs its head happily to the music and does one small proud wing flap. No objects fall or move off the table. Premium mobile ad lighting, crisp textures. Audio: playful kid pop music, soft cheerful chirp SFX.

Cartoon green plush bird with a chef hat, round body with no legs, sitting on a wooden kitchen table next to a neat stack of golden pancakes with syrup. Static camera, eye-level, shallow depth of field. The bird bobs its head happily to the music and does one small proud wing flap. No objects fall or move off the table. Premium mobile ad lighting, crisp textures. Audio: playful kid pop music, soft cheerful chirp SFX.

Output

Energy drink urban night · text-to-video · 9:16 · 1080p · 5s · music + SFX

Example prompt

Vertical cinematic ad. Fictional white teenage skater with freckles and curly red hair, yellow rain jacket, opens a slim silver energy can on a wet neon Berlin side street at night. Rain droplets on lens, reflections on pavement, static camera at waist height. They take one sip and nod to the beat. Audio: trap-pop instrumental, rain ambience, crisp can open SFX.

Vertical cinematic ad. Fictional white teenage skater with freckles and curly red hair, yellow rain jacket, opens a slim silver energy can on a wet neon Berlin side street at night. Rain droplets on lens, reflections on pavement, static camera at waist height. They take one sip and nod to the beat. Audio: trap-pop instrumental, rain ambience, crisp can open SFX.

Output

SaaS promo with supers · text-to-video · 16:9 · 720p · 8s · instrumental

Example prompt

Modern glass office at golden hour, empty desk with glowing laptop showing an abstract purple dashboard UI (no readable brand logos). Large clean white text overlay centered: "SHIP 10× FASTER". Camera slow push-in toward laptop. Minimal motion graphics sparkle on the text. Audio: upbeat electronic corporate instrumental track, no voiceover.

Modern glass office at golden hour, empty desk with glowing laptop showing an abstract purple dashboard UI (no readable brand logos). Large clean white text overlay centered: "SHIP 10× FASTER". Camera slow push-in toward laptop. Minimal motion graphics sparkle on the text. Audio: upbeat electronic corporate instrumental track, no voiceover.

Output

Kitchen appliance hero · text-to-video · 16:9 · 720p · 5s · silent

Example prompt

Matte black air fryer on white quartz kitchen counter, morning side light, no people. Lid opens slowly revealing golden fries inside, subtle steam. Static camera, Apple-style product cleanliness. No music, no voice.

Matte black air fryer on white quartz kitchen counter, morning side light, no people. Lid opens slowly revealing golden fries inside, subtle steam. Static camera, Apple-style product cleanliness. No music, no voice.

Output

AI music clip · image-to-video + audio · 9:16 · 1080p

Example prompt

Performance shot: artist sings to camera with natural lip sync, subtle head and shoulder movement, handheld music-video energy but stable face framing, stage lights pulse gently. Lyrics she sings (original): "Neon veins under my skin, pull me back and pull me in. Every pulse is telling me to start again — start again, start again." Audio: upbeat electro-pop beat, female vocal singing the lyrics above clearly, light reverb, concert crowd ambience.

Performance shot: artist sings to camera with natural lip sync, subtle head and shoulder movement, handheld music-video energy but stable face framing, stage lights pulse gently. Lyrics she sings (original): "Neon veins under my skin, pull me back and pull me in. Every pulse is telling me to start again — start again, start again." Audio: upbeat electro-pop beat, female vocal singing the lyrics above clearly, light reverb, concert crowd ambience.

Output

Rap lyric visualizer · text-to-video · 9:16 · 720p · 8s

Example prompt

Fictional Black rapper in oversized hoodie performs in a parking garage at night, purple LED tubes, light fog, camera slow orbit at chest height, confident hand gestures on beat. On-screen graffiti-style text overlay: "MIDNIGHT DROP". Lyrics he raps (original): "Midnight drop — yeah we run the block. Neon signs and the bass won't stop. Started from the bottom of the parking lot, now every speaker hit every spot. Yeah! Let's go!" Audio: trap beat with heavy 808s, rapper delivers the lyrics above with clear enunciation, ad-libs on the last line.

Fictional Black rapper in oversized hoodie performs in a parking garage at night, purple LED tubes, light fog, camera slow orbit at chest height, confident hand gestures on beat. On-screen graffiti-style text overlay: "MIDNIGHT DROP". Lyrics he raps (original): "Midnight drop — yeah we run the block. Neon signs and the bass won't stop. Started from the bottom of the parking lot, now every speaker hit every spot. Yeah! Let's go!" Audio: trap beat with heavy 808s, rapper delivers the lyrics above with clear enunciation, ad-libs on the last line.

Source image Acoustic session source still

Output

Acoustic session · image-to-video · 16:9 · 720p · lyrics

Example prompt

Intimate live session: she strums acoustic guitar and sings with accurate lip sync, minimal body sway, static camera, warm film grain. Lyrics she sings (original): "City lights are fading slow, I packed my bags and let you go. Every street remembers us, but I'm not asking for your trust." Audio: fingerpicked acoustic guitar plus her voice singing the lyrics above, warm studio room tone, no backing track.

Intimate live session: she strums acoustic guitar and sings with accurate lip sync, minimal body sway, static camera, warm film grain. Lyrics she sings (original): "City lights are fading slow, I packed my bags and let you go. Every street remembers us, but I'm not asking for your trust." Audio: fingerpicked acoustic guitar plus her voice singing the lyrics above, warm studio room tone, no backing track.

Output

Medieval knight entrance · text-to-video · 16:9 · 1080p · 24 fps · 6s

Example prompt

Armored knight in weathered silver armor walks through a muddy medieval village toward a stone castle gate, torches flicker, villagers scatter. Slow tracking camera behind the knight, Unreal Engine 5 look, volumetric fog, late afternoon sun. Audio: epic orchestral swell, distant blacksmith hammer, boots on cobblestone.

Armored knight in weathered silver armor walks through a muddy medieval village toward a stone castle gate, torches flicker, villagers scatter. Slow tracking camera behind the knight, Unreal Engine 5 look, volumetric fog, late afternoon sun. Audio: epic orchestral swell, distant blacksmith hammer, boots on cobblestone.

Source image Boss reveal source still

Output

Boss reveal · image-to-video · 16:9 · 1080p · 4s · silent

Example prompt

Slow dramatic push-in toward the golem, cracks glow brighter, embers float upward, camera stable, cinematic game announcement trailer energy. No audio.

Slow dramatic push-in toward the golem, cracks glow brighter, embers float upward, camera stable, cinematic game announcement trailer energy. No audio.

Output

Cozy farming sim · text-to-video · 1:1 · 720p · 5s

Example prompt

Round fox farmer character waters sunflowers in a pastel meadow, butterflies drift, static isometric-leaning camera, soft morning light, Animal Crossing mood. Audio: gentle ukulele and birdsong.

Round fox farmer character waters sunflowers in a pastel meadow, butterflies drift, static isometric-leaning camera, soft morning light, Animal Crossing mood. Audio: gentle ukulele and birdsong.

Output

Spaceship hangar fly-through · text-to-video · 16:9 · 1080p · 8s

Example prompt

Camera slowly flies forward through a massive spaceship hangar, fighters on lifts, orange welding sparks, crew silhouettes, god rays through smoke, photoreal AAA sci-fi, 16:9. Audio: industrial hum, distant PA announcement (unintelligible), cinematic percussion build.

Camera slowly flies forward through a massive spaceship hangar, fighters on lifts, orange welding sparks, crew silhouettes, god rays through smoke, photoreal AAA sci-fi, 16:9. Audio: industrial hum, distant PA announcement (unintelligible), cinematic percussion build.

Source image On-model fashion source still

Output

On-model fashion loop · image-to-video · 9:16 · 720p · 5s

Example prompt

Subtle fashion editorial movement, slight playful movements, model maintains confident presence, natural breathing motion, gentle shoulder shift, soft micro head movement, static camera, premium ecommerce showcase, controlled minimal motion. Audio: minimal pop groove, no vocals.

Subtle fashion editorial movement, slight playful movements, model maintains confident presence, natural breathing motion, gentle shoulder shift, soft micro head movement, static camera, premium ecommerce showcase, controlled minimal motion. Audio: minimal pop groove, no vocals.

Source image Jewelry source still

Output

Jewelry macro · image-to-video · 2:3 · 1080p · 4s

Example prompt

Slow macro push-in, pendant catches moving light reflections, velvet texture stable, static tripod, luxury boutique atmosphere. Audio: subtle ambient piano, very quiet.

Slow macro push-in, pendant catches moving light reflections, velvet texture stable, static tripod, luxury boutique atmosphere. Audio: subtle ambient piano, very quiet.

Source image Fashion flash sale source still

Output

Fashion flash sale · image-to-video · 1:1 · 720p · 4s · silent

Example prompt

Minimal fashion motion: gentle weight shift, coat fabric subtly moves, model maintains eye contact with camera, static tripod shot, premium lookbook energy. No speech, no music.

Minimal fashion motion: gentle weight shift, coat fabric subtly moves, model maintains eye contact with camera, static tripod shot, premium lookbook energy. No speech, no music.

Source image Sneaker source still

Output

Sneaker turntable · image-to-video · 1:1 · 1080p · 4s · silent

Example prompt

Slow 15-degree product rotation on invisible turntable, sneaker stays sharp and centered, static camera, premium SNKRS-style product video, no hands, no text.

Slow 15-degree product rotation on invisible turntable, sneaker stays sharp and centered, static camera, premium SNKRS-style product video, no hands, no text.

Output

Podcast dialogue · text-to-video · 16:9 · 720p · 10s · two speakers

Example prompt

Two fictional French men speaking English with a clearly noticeable French accent, natural conversational tone, relaxed pacing, discussing a new AI video generation model, intelligent but casual tech founders vibe. They remain seated on a podcast sofa, facing slightly toward each other. Subtle conversational movement only: light head nods, very small hand movements, natural breathing. Static camera, realistic podcast studio, documentary realism. Dialogue — Man 1: "Have you tested P-Video-2 yet? The speed-to-quality cost ratio is impressive." Man 2: "I did! I made a ten-second video in ten seconds for only ten cents. I really love it." Audio: warm podcast studio room tone, faint lo-fi intro music bed under the conversation at low volume, natural dialogue levels — not dry voice-only.

Two fictional French men speaking English with a clearly noticeable French accent, natural conversational tone, relaxed pacing, discussing a new AI video generation model, intelligent but casual tech founders vibe. They remain seated on a podcast sofa, facing slightly toward each other. Subtle conversational movement only: light head nods, very small hand movements, natural breathing. Static camera, realistic podcast studio, documentary realism. Dialogue — Man 1: "Have you tested P-Video-2 yet? The speed-to-quality cost ratio is impressive." Man 2: "I did! I made a ten-second video in ten seconds for only ten cents. I really love it." Audio: warm podcast studio room tone, faint lo-fi intro music bed under the conversation at low volume, natural dialogue levels — not dry voice-only.

Output

Employer brand sizzle · text-to-video · 16:9 · 1080p · 6s · music only

Example prompt

Diverse team collaborating in a bright modern office — handshake, laptop stand-up, high-five near a window with city view. Slow cinematic push-in through the open-plan space, warm natural light, documentary corporate film look, no readable company logos. Audio: uplifting indie corporate instrumental track, medium energy, no voiceover, no dialogue.

Diverse team collaborating in a bright modern office — handshake, laptop stand-up, high-five near a window with city view. Slow cinematic push-in through the open-plan space, warm natural light, documentary corporate film look, no readable company logos. Audio: uplifting indie corporate instrumental track, medium energy, no voiceover, no dialogue.

Output

CEO product update · text-to-video · 16:9 · 720p · 6s · voice + music bed

Example prompt

Fictional Black woman CEO in grey suit stands in bright modern HQ lobby, static camera medium shot. She speaks confidently: "We're shipping the integration this quarter." Clean lower-third text overlay: "Q3 ROADMAP UPDATE". Subtle smile, professional lighting. Audio: her voice clear and forward, subtle news-style corporate music bed underneath at low volume, very light office ambience — never voice-only.

Fictional Black woman CEO in grey suit stands in bright modern HQ lobby, static camera medium shot. She speaks confidently: "We're shipping the integration this quarter." Clean lower-third text overlay: "Q3 ROADMAP UPDATE". Subtle smile, professional lighting. Audio: her voice clear and forward, subtle news-style corporate music bed underneath at low volume, very light office ambience — never voice-only.

Integration

P-Video-2 uses the same Pruna predictions API as other performance models. Text-to-video needs only a prompt; image-to-video and audio-conditioned runs upload files first.

Tip

For more information on how to use the API, see the API Reference.

API endpoint: Base URL: https://api.pruna.ai/v1/predictions

Authentication

-H 'apikey: YOUR_API_KEY'
-H 'Model: p-video-2'

Text-to-video (asynchronous)

curl -X POST 'https://api.pruna.ai/v1/predictions' \
  -H 'Content-Type: application/json' \
  -H 'apikey: YOUR_API_KEY' \
  -H 'Model: p-video-2' \
  -d '{
    "input": {
      "prompt": "A sports car drifting through a neon-lit city at night, cinematic aerial shot",
      "duration": 5,
      "resolution": "720p",
      "aspect_ratio": "16:9"
    }
  }'

Text-to-video (synchronous)

curl -X POST 'https://api.pruna.ai/v1/predictions' \
  -H 'Content-Type: application/json' \
  -H 'apikey: YOUR_API_KEY' \
  -H 'Model: p-video-2' \
  -H 'Try-Sync: true' \
  -d '{
    "input": {
      "prompt": "A sports car drifting through a neon-lit city at night, cinematic aerial shot",
      "duration": 5,
      "resolution": "720p",
      "draft": true
    }
  }'

Image-to-video

Upload a reference image, then pass its file URL:

curl -X POST "https://api.pruna.ai/v1/files" \
  -H "apikey: YOUR_API_KEY" \
  -F "content=@/path/to/your/file.jpg"

curl -X POST 'https://api.pruna.ai/v1/predictions' \
  -H 'Content-Type: application/json' \
  -H 'apikey: YOUR_API_KEY' \
  -H 'Model: p-video-2' \
  -d '{
    "input": {
      "prompt": "The camera slowly pushes in, the person turns their head and smiles",
      "image": "https://api.pruna.ai/v1/files/fqadqq42xq",
      "duration": 5,
      "resolution": "720p"
    }
  }'

Configuration

Required parameters

Parameter	Type	Description
`prompt`	string	Text prompt for video generation

Optional parameters

Parameter	Type	Default	Description
`duration`	integer	—	Duration in seconds (1–20). Leave empty to let the model choose from the prompt. Ignored when `audio` is provided
`image`	string (URI)	—	Input image for image-to-video (jpg, jpeg, png, webp)
`audio`	string (URI)	—	Input audio to condition generation (flac, mp3, wav)
`resolution`	string	`"720p"`	`"720p"` or `"1080p"`
`fps`	integer	`24`	`24` or `48`
`aspect_ratio`	string	`"16:9"`	`"16:9"`, `"9:16"`, `"4:3"`, `"3:4"`, `"3:2"`, `"2:3"`, `"1:1"`. Ignored when an input image is provided
`seed`	integer	random	Random seed for reproducible generation
`draft`	boolean	`false`	Faster, lower-quality preview. Billed at half the standard per-second rate
`save_audio`	boolean	`true`	Save the video with audio
`last_frame_image`	string (URI)	—	Reference image for the last frame
`prompt_upsampling`	boolean	`true`	Enhance the prompt
`disable_safety_filter`	boolean	`true`	Disable safety filter for prompts and input images

Argument recommendations

Use these patterns for consistent quality:

prompt: follow subject / action / scene, then optional camera, lighting, style, and audio. For image-to-video, keep motion subtle and aligned with the reference frame.
image: high-quality, well-lit still; aspect_ratio is ignored when an image is set. Optional last_frame_image for end-frame control.
audio: use imported audio when you need a specific track; duration follows the audio and duration is ignored. Lead lip-sync demos with native speech (save_audio: true, no imported audio).
duration: 1–20 seconds, or leave empty so the model chooses from the prompt.
resolution / fps: iterate in 720p with draft: true, then rerun finals with draft: false. Prefer 24 fps at 1080p for stability; use 48 fps at 720p for smoother action.
prompt_upsampling: leave true for production; compare true vs false on the same seed before scaling.