The old website stays available at og.deepany.ai but no longer receives updates and may stop working over time. Your account and credits work on both, and you can download your creations from it anytime. Open the old website

The example collection

Look closer.
Imagine further.

Inspect real results and the ingredients behind them.

Results and inputs are different. A recorded recipe includes the creative settings and original inputs. A reference image alone is never labeled as a generated video.

30 examples

Text to Image / Verified Studio coastal editorialRecorded Studio result
DeepAny Studio image output showing a woman in a deep-red dress on a windswept coast
Actual Studio-generated result. Its matching prompt, settings and seed can be opened from the library.

Images / Studio result

Chase the feeling.

A Studio-generated coastal editorial with a deep-red dress, tall grass, sea breeze and warm sunset light.

Example direction

A windswept coast with a flowing deep-red dress in tall grass above the sea at sunset. Natural fabric movement, warm evening light, cinematic editorial photography, realistic detail, clean composition, no text, no watermark.

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Image
Image style
Realistic
Aspect ratio
1:1
Quality
Detailed · 8 sampling steps
Seed
1623251206
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Image direction / Product studyRecorded Studio result
Fragrance bottle on a sculptural pedestal
Actual Studio-generated result. Its matching prompt, settings and seed can be opened from the library.

Images / Studio result

Objects with presence.

A Studio-generated clear perfume bottle on a dark stone plinth in soft window light.

Example direction

A sculptural perfume bottle on a dark stone plinth, soft window light, quiet shadows, editorial product photography, refined glass and metal details, premium studio styling, realistic detail, clean composition, no text, no watermark.

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Image
Image style
Realistic
Aspect ratio
1:1
Quality
Detailed · 8 sampling steps
Seed
4141089598
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Image direction / ArchitectureRecorded Studio result
Modern villa with warm interior light at dusk
Actual Studio-generated result. Its matching prompt, settings and seed can be opened from the library.

Images / Studio result

Architecture, imagined.

A saved DeepAny Studio Text to Image result: a glass-and-concrete coastal house with warm interior lights at dusk.

Example direction

A modern glass-and-concrete coastal house on a rocky hillside, warm interior lights at dusk, architectural photography, clean geometry, natural materials, refined editorial composition, no people, no text, no watermark.

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Image
Image style
Realistic
Aspect ratio
1:1
Quality
Detailed · 8 sampling steps
Seed
2533448545
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Composition / Landscape studyRecorded Studio result
Astronaut standing in a bright desert landscape
Actual Studio-generated result. Its matching prompt, settings and seed can be opened from the library.

Images / Studio result

Somewhere, beyond.

A saved DeepAny Studio Text to Image result: an astronaut standing in a wide red desert landscape.

Example direction

A lone astronaut standing on a windswept red desert beneath a pale blue sky, cinematic landscape photography, crisp silhouette, wide environmental composition, realistic textures, no text, no watermark.

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Image
Image style
Realistic
Aspect ratio
1:1
Quality
Detailed · 8 sampling steps
Seed
1089539455
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / 1080p woodland motionRecorded Studio result
Actual five-second Studio result at 1920 × 1080 and 24 fps. Sunlit ferns move gently in a woodland scene. Open the recorded recipe to repeat the workflow.

Video / Studio result

A quiet woodland, brought to life.

A five-second 1080p Studio woodland result. Its exact final prompt, enhancement settings and fixed seed are included.

Example direction

integrated_multimodal_description: [Shot 1] A steady wide shot of a quiet green woodland at sunrise. Detailed ferns fill the foreground, with tall tree trunks behind them. Warm golden shafts of morning light fall diagonally between the trees, illuminating thin mist and leaves. Leaves and fern fronds sway gently and continuously in a light breeze throughout the five-second clip. The camera stays completely static with no cuts, pans, tilts or zooms. No people, speech, text or watermark. overall_soundscape: Soft natural birdsong with gentle rustling leaves and fern fronds in the breeze. No speech or artificial sounds. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Video Generator
Size
1920 × 1080 · 16:9
Duration
5 seconds · 24 fps
Image, video and upscale enhancement
On · 50% each
Motion control
50%
Sampling
Euler · workflow default schedule
Seed
2592251842
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Guided Edit / Interior studyRecorded Studio result
Studio-generated room with a terracotta armchair
Studio edit with navy blue woven upholstery
BeforeAfter
Actual Studio source and edit. The terracotta upholstery becomes navy woven fabric; the sofa and overall room composition remain recognizable. Review small lighting and texture differences as well as the intended recolor.

Editing / Studio result

Change the conversation.

Studio changed the terracotta armchair upholstery to navy blue woven fabric. The original image and settings are included in the recipe.

Example direction

Change only the terracotta armchair upholstery to deep navy blue woven fabric. Preserve the chair shape, stitching, cushions and position. Keep the cream sofa, walls, oak floor, window, camera view and natural daylight unchanged. Match the existing shadows. No text, no watermark.

Recorded prompt.

Recipe, inputs & settings
Workflow
Guided Edit
Edit mode
Standard
Quality
Detailed · 16 sampling steps
Seed
1082279553
Banner
Removed
Source input
Recorded source input for Give the chair a new material Download source input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Inpaint Anything / Red boatRecorded Studio result
Original synthetic lake scene with a natural wooden boat
Native Inpaint result with a bright red boat and a dark reflection
BeforeAfter
Actual native Studio recolor. The wooden hull becomes bright red while the lake, mountains and framing remain recognizable. The reflected underside stays very dark; inspect both sides with the comparison control.

Editing / Studio result

Room for the scene.

A saved DeepAny Studio Inpaint Anything result: the selected wooden boat is recolored bright red while the lake, mountains, and reflections remain in context.

Example direction

A bright red painted wooden rowboat on a calm alpine lake, vivid scarlet red hull, realistic paint texture and natural dawn light. Its reflection below the boat has matching dark red color on the lake surface.

Recorded prompt.

Recipe, inputs & settings
Workflow
Inpaint Anything
Region
boat. reflection of the boat.
Quality
Detailed · 20 sampling steps
Original-depth conditioning
On for recoloring
Seed
Worker-randomized; exact pixel reproduction is not guaranteed
Banner
Removed
Source input
Recorded source input for Recolor a boat in its landscape Download source input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Face Swap / Identity studyRecorded Studio result
Synthetic target portrait before Face Swap
Studio Face Swap result with the source identity and original target scene
BeforeAfter
Actual Studio Face Swap using two synthetic portraits. Compare the target and result, then open the recipe to view the separate identity source and detailed controls.

Identity / Studio result

A face, in context.

A detailed Studio Face Swap using two synthetic portraits. The exact identity source, target and generation controls are included in its recipe.

Example direction

Review the authorized face replacement in the supplied target. Check eyes, mouth, expression, hairline, lighting, and boundaries. For video, inspect head turns and partial occlusion.

Review checklist only; Face Swap is driven by the recorded inputs and controls below, not this text.

Recipe, inputs & settings
Workflow
Photo Face Swap
Inputs
Two synthetic Studio portraits
Target face
First detected face · index 0
Swap mode
Experimental
Face detail
512 × 512
Enhancement
High detail · 50% blend
Boundary
Soft boundary on; obstruction detection off
Additional detectors and rotations
Off
Age changes
Off
Banner
Removed
Seed
Not applicable to this workflow
Identity input
Recorded identity input for A review brief for identity work Download identity input
Target input
Recorded target input for A review brief for identity work Download target input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Talking Face / Photo + audioRecorded Studio result
Actual five-second Studio result from a synthetic portrait and an original generated recording. Both source files are included in the recipe. Compare the mouth, eyes and timing with the source audio; the requested speech script has not been independently transcribed.

Video / Studio result

Photo meets performance.

Actual Talking Face output with its synthetic portrait, original speech recording and saved generation recipe.

Example direction

Frame alignment: Picture 1 is the exact opening portrait at zero seconds. integrated_multimodal_description: [Shot 1] The original fictional adult woman in the supplied portrait faces the camera in the same softly lit room. Preserve her face, hair, clothes and background throughout a five-second chest-up shot. Her lips and jaw follow the supplied original speech recording naturally, with restrained head movement, subtle blinking and a calm friendly expression. Keep the camera locked and finish with a natural settled expression. No cuts, new people, text or watermark. overall_soundscape: Reuse the supplied audio performance unchanged. Do not add voices or sound effects. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Talking Face
Inputs
Synthetic portrait + original generated speech
Output
1080 × 1080 · 5 seconds · 24 fps
Enhancement
Video and upscale · 50%
Sampling
Euler · workflow default schedule
Seed
2026091308
Banner
Removed
Consent
Confirm for your intended use in Studio
Portrait input
Recorded portrait input for A photo and an approved voice Download portrait input
Audio input
Download audio input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Text to Audio / Recorded Studio resultRecorded Studio result

Listen to the morning breeze.

Actual five-second Studio recording generated from the birdsong and rustling-leaves prompt below. Use the player to judge the sound; open its recipe for the exact duration, sampler and seed.

Audio / Studio result

A sound to listen for.

A saved DeepAny Studio Text to Audio result with soft birdsong and leaves rustling in a gentle morning breeze.

Example direction

Soft birdsong and leaves rustling in a gentle morning breeze. No speech, no music.

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Audio
Format
Lossless FLAC · stereo · 32 kHz · 16 bit
Duration
5 seconds
Sampling
Euler · workflow default schedule
Seed
2026091302
Source
Text only; no imported audio

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Text to Audio / Recorded Studio resultRecorded Studio result

One calm voice, from text.

Read transcript

Requested script: "Every great project begins with a single idea. Today we are going to turn that idea into something you can see, and something you can hear."

Actual ten-second Text to Audio recording from the spoken-script prompt below. The voice is synthetic; no real speaker was recorded or cloned. The same file is the source for the Voice Cloning and Audio Extender examples.

Audio / Studio result

One calm voice, from text.

Actual DeepAny Studio Text to Audio result: ten seconds of calm synthetic speech from a written script. No real speaker was recorded or cloned.

Example direction

A calm adult man says clearly in English: "Every great project begins with a single idea. Today we are going to turn that idea into something you can see, and something you can hear." Natural conversational delivery, measured pace, clean ending, quiet recording room. One speaker only. No music, no other voices, no sound effects. Speak only the quoted sentences, not these instructions.

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Audio
Format
Lossless FLAC · stereo · 32 kHz · 16 bit
Duration
10 seconds
Audio quality
Very high
Seed
1917089614
Voice
Generated from text; no cloned identity

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Voice Cloning / Recorded Studio resultRecorded Studio result

A new line, in the same voice.

Read transcript

Requested words: "Every great project begins with a single idea. This new line is spoken in the same voice, without recording anything again."

Actual ten-second Audio & Voice Cloning result. The voice reference was a synthetic Studio Text to Audio recording, not a real person, so no one's voice was cloned. Open the recipe to hear that reference next to this result.

Audio / Studio result

A new line, in the same voice.

Actual DeepAny Studio Audio & Voice Cloning result: new words in the voice of the synthetic Text to Audio recording, which is included in the recipe as the voice reference.

Example direction

Every great project begins with a single idea. This new line is spoken in the same voice, without recording anything again.

Recorded prompt.

Recipe, inputs & settings
Workflow
Audio & Voice Cloning
Voice reference
Synthetic Studio Text to Audio recording · 10 seconds
Format
Lossless FLAC · stereo · 32 kHz · 16 bit
Duration
10 seconds
Spoken language
English
Seed
739244868
Voice input
Download voice input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Audio Extender / Recorded Studio resultRecorded Studio result

Ten seconds in, ten more out.

Read transcript

0:00 to 0:10, source: "Every great project begins with a single idea. Today we are going to turn that idea into something you can see, and something you can hear." Join at 0:10. 0:10 to 0:20, new continuation in the same voice, directed to close the thought "Start with the idea, then let the work show how far it can go."

Actual twenty-second Audio & Voice Extender result. 0:00 to 0:10 is the source recording, unchanged apart from a short blend just before the join; the join is at 0:10 and everything after it is new continuation in the same voice. Open the recipe to hear the source on its own.

Audio / Studio result

Ten seconds in, ten more out.

Actual DeepAny Studio Audio & Voice Extender result: the ten-second synthetic speech source, then ten seconds of new continuation in the same voice. The join is at 0:10.

Example direction

Continue the same calm explanation in the same voice and natural pace, closing the thought: "Start with the idea, then let the work show how far it can go."

Recorded prompt.

Recipe, inputs & settings
Workflow
Audio & Voice Extender
Source
Synthetic Studio Text to Audio recording · 10 seconds
Extension
10 seconds from the end · join at 0:10
Format
Lossless FLAC · stereo · 32 kHz · 16 bit · 20 seconds
Seed
Not recorded for this example
Source input
Download source input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Edit / Balanced EditRecorded Studio result
Source video
Edited video
The cyclist starts in orange. Balanced Edit changes the shirt to a vivid cobalt blue cycling jacket while keeping the rider, bicycle, coastal road, lighting, and camera motion coherent.

Video / Studio result

Change the look. Keep the ride.

A GPU-verified Balanced Edit changes the cyclist from orange to cobalt blue while keeping the bicycle, coastal road, and camera motion coherent.

Example direction

Change the cyclist orange shirt into a vivid cobalt blue cycling jacket. Preserve the rider identity, bicycle, road, coastal background, camera movement, lighting, and natural cycling motion.

Recorded prompt.

Recipe, inputs & settings
Workflow
Video Edit
Mode
Balanced Edit
Input
5-second cyclist source range
Output
864 × 480 · 5 seconds · 24 fps
Edit strength
65%
Seed
26091501
Acceptance
GPU-verified on signed backend
Source input
Download source input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Edit / Preserve MotionRecorded Studio result
Source video
Edited video
Preserve Motion changes the cyclist clothing to cobalt blue while leaning more heavily on the encoded source clip for timing, pose progression, bicycle motion, and camera movement.

Video / Studio result

Hold onto the motion.

A GPU-verified Preserve Motion result keeps the cyclist motion and framing close to the source while carrying the cobalt-blue clothing edit through the clip.

Example direction

Change the cyclist orange shirt into a vivid cobalt blue cycling jacket. Preserve the rider identity, bicycle, road, coastal background, camera movement, lighting, and natural cycling motion.

Recorded prompt.

Recipe, inputs & settings
Workflow
Video Edit
Mode
Preserve Motion
Input
5-second cyclist source range
Output
864 × 480 · 5 seconds · 24 fps
Edit strength
65%
Seed
26091502
Acceptance
GPU-verified on signed backend
Source input
Download source input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Edit / Timed EditRecorded Studio result
Source video
Edited video
Timed Edit keeps the original orange-shirt frames before 2.5 seconds, then starts the cobalt-blue edited continuation at the chosen change point.

Video / Studio result

Change it at the right moment.

A GPU-verified Timed Edit leaves the orange-shirt source unchanged before 2.5 seconds, then begins the cobalt-blue edited continuation.

Example direction

Change the cyclist orange shirt into a vivid cobalt blue cycling jacket. Preserve the rider identity, bicycle, road, coastal background, camera movement, lighting, and natural cycling motion.

Recorded prompt.

Recipe, inputs & settings
Workflow
Video Edit
Mode
Timed Edit
Input
5-second cyclist source range
Output
864 × 480 · 5 seconds · 24 fps
Edit strength
65%
Change time
2.5 seconds
Seed
26091503
Acceptance
GPU-verified on signed backend
Source input
Download source input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Text to Video / Coastal trackingRecorded Studio result
An actual five-second Studio shot at 1920 × 1080. The cyclist continues pedaling while the side-tracking camera follows. The complete recorded direction and controls are available. Small roadside fixtures can vary from the requested exclusions.

Video / Studio result

A camera that follows the idea.

A five-second 1080p Studio result: a cyclist tracked along a coastal road. The exact prompt, settings and seed are included in the recipe.

Example direction

integrated_multimodal_description: [Shot 1] A single adult cyclist in a plain muted orange jersey, dark shorts and a bicycle helmet rides a simple black road bicycle along a quiet coastal road in soft morning light. The calm sea and low coastal hills remain visible beyond a simple roadside barrier. The camera tracks smoothly alongside the cyclist at a steady distance for five seconds, maintaining a clear side view of the rider and complete bicycle. Continuous natural pedaling, both hands on the handlebars, gently moving fabric, stable bicycle wheels and road geometry. Continue riding forward toward the open road at the end, without stopping or cutting. No other vehicles, people, signs, lettering, logos or watermark. overall_soundscape: Soft coastal wind and quiet rolling bicycle tires. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Video
Output
1920 × 1080 · 5 seconds · 24 fps
Enhancement
Image, video and upscale · 50%
Sampling
Euler · workflow default schedule
Seed
2026091309
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Image to Video / Product motionRecorded Studio result
Actual five-second Studio video generated from the perfume still at 1080 × 1080. The slow push approaches the bottle without adding labels or hands. Open its exact source and settings in the recipe.

Video / Studio result

Movement around a still subject.

A recorded perfume still becomes a slow product push-in. The first-frame image and settings are included in the recipe.

Example direction

integrated_multimodal_description: [Shot 1] Start exactly from the supplied image: a clear perfume bottle with a black cap and pale golden liquid stands on a dark stone plinth beside a softly lit window. A slow, restrained camera push approaches the bottle over five seconds. The bottle and plinth remain still. Glass reflections shift subtly and naturally with the camera. Preserve the bottle shape, cap, liquid and room throughout the shot. No cuts, hands, extra objects, label, lettering, logos or watermark. overall_soundscape: Quiet indoor room tone. No speech or sound effects. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Image to Video
Size
1080 × 1080 · 1:1
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Sampling
Euler · workflow default schedule
Seed
2026091306
Banner
Removed
Startframe input
Recorded startFrame input for Movement around a still subject Download startFrame input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Extender / Coastal continuationRecorded Studio result
The actual combined Studio output: five seconds of the source coastal ride, followed by five new seconds. The cyclist and camera keep moving forward. Review the transition around five seconds and compare it with the downloadable source.

Video / Studio result

The next beat, not a new scene.

The original five-second coastal ride followed by five newly generated seconds. Inspect the join at five seconds.

Example direction

Frame alignment: Picture 1 is the extracted ending frame of the source clip and the opening frame of this five-second continuation. integrated_multimodal_description: [Shot 1] Continue the same adult cyclist in an orange jersey, dark shorts and helmet riding the same black bicycle forward along the coastal road. Match the supplied ending frame, the calm sea, distant coastline, roadside barrier and soft morning light. Continue smooth natural pedaling and the existing side-tracking camera at a steady distance, retaining the same rider, bicycle, clothing and direction of travel. No cut, reset, stop, new people, lettering or watermark. overall_soundscape: Soft coastal wind and quiet rolling bicycle tires. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Video Extender
Input
5-second recorded coastal video
Added duration
5 seconds
Combined output
1920 × 1080 · approximately 10 seconds · 24 fps
Enhancement
Video and upscale · 50%
Sampling
Euler · workflow default schedule
Seed
2026091313
Banner
Removed
Source input
Download source input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / Portrait motionRecorded Studio result
Actual five-second Studio generation at 1920 × 1080. The original fictional subject makes subtle movements in soft window light as the camera approaches. The exact recorded prompt and settings are available.

Video / Studio result

A restrained performance.

A five-second 1080p portrait with subtle head and eye movement and a gentle camera approach.

Example direction

integrated_multimodal_description: [Shot 1] An original fictional adult woman with short dark hair, wearing a simple charcoal crew-neck top, framed from the chest up beside a softly lit window in a quiet neutral studio. She turns gently toward the camera in soft window light. A slow restrained camera push-in approaches her face over five seconds. The background remains stable. Natural pauses and one subtle blink; finish on a calm held expression without a cut. Natural skin texture, consistent face and clothing. No text or watermark. overall_soundscape: Quiet indoor room tone. No speech or sound effects. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Video Generator
Output
1920 × 1080 · 5 seconds · 24 fps
Enhancement
Image, video and upscale · 50%
Sampling
Euler · workflow default schedule
Seed
2026091307
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / A whole day, in five secondsRecorded Studio result
Actual DeepAny Studio start-and-end-frame result. The first frame is a dawn skyline and the last is the same view at night; the video fills the five seconds between them with moving clouds, shifting light and traffic.

Video / Studio result

A whole day, in five seconds.

Actual DeepAny Studio start-and-end-frame result. The first frame is a dawn skyline and the last is the same view at night; the video fills the five seconds between them with moving clouds, shifting light and traffic.

Example direction

Frame alignment: Picture 1 is the opening frame at zero seconds and Picture 2 is the final frame at five seconds. integrated_multimodal_description: [Shot 1] A sweeping time-lapse aerial shot over a coastal city, starting at pink dawn and racing through the day into a glittering night. Clouds stream across the sky, the sun arcs and sets, shadows sweep across the towers, windows light up floor by floor, and long streaks of car headlights flow along the roads and across the bridge. The camera glides steadily forward and slightly down over the bay with strong parallax, ending on the lit night skyline. Continuous motion in every part of the frame. No text, no logos, no watermark. overall_soundscape: A soft rising city ambience with distant traffic and a gentle whoosh of passing time. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Start and end frames
Size
1080 × 1080 · 1:1
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed
Startframe input
Recorded startFrame input for A whole day, in five seconds Download startFrame input
Lastframe input
Recorded lastFrame input for A whole day, in five seconds Download lastFrame input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / The pour, from first drop to latte artRecorded Studio result
Actual DeepAny Studio start-and-end-frame result. The first frame is a black espresso and the last frame is the finished latte; the video shows the pour in between.

Video / Studio result

The pour, from first drop to latte art.

Actual DeepAny Studio start-and-end-frame result. The first frame is a black espresso and the last frame is the finished latte; the video shows the pour in between.

Example direction

Frame alignment: Picture 1 is the opening frame at zero seconds and Picture 2 is the final frame at five seconds. integrated_multimodal_description: [Shot 1] A macro cafe shot of a barista pouring steamed milk into an espresso. The milk stream arcs from the pitcher, hits the dark crema and blooms into white swirls, the pitcher rocks side to side and the pattern folds into a perfect leaf rosette, then the pitcher lifts away. Steam curls upward and drifts, tiny crema ripples spread across the surface, and the camera makes a slow, smooth push-in and downward tilt over the cup, ending on the finished latte art. Continuous fluid motion. No text, logos or watermark. overall_soundscape: Gentle milk pouring and a soft cafe room tone with a faint steam hiss. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Start and end frames
Size
1080 × 1080 · 1:1
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed
Startframe input
Recorded startFrame input for The pour, from first drop to latte art Download startFrame input
Lastframe input
Recorded lastFrame input for The pour, from first drop to latte art Download lastFrame input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / A camera that can keep upRecorded Studio result
Actual DeepAny Studio Text to Video result: a five-second 21:9 tracking shot along a coastal road, from a single text prompt.

Video / Studio result

A camera that can keep up.

Actual DeepAny Studio Text to Video result: a five-second 21:9 tracking shot along a coastal road, from a single text prompt.

Example direction

integrated_multimodal_description: [Shot 1] A fast, dramatic cinematic tracking shot along a rugged sea-cliff road at golden hour. A rider in a dark jacket on a red motorcycle accelerates along the road toward the camera, then rips past it as the camera whips around in a smooth arc to follow from behind. Dust and grit kick up from the road, wind flattens the tall grass, waves crash against the rocks below with spray drifting through the light, clouds race across the sky. Strong parallax and continuous motion in every part of the frame, with smooth, confident camera movement for the full five seconds. No text, no logos, no watermark. overall_soundscape: Rushing wind, crashing waves, a deep engine roar rising as it passes the camera and fading away. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Video
Size
2520 × 1080 · 21:9
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / Drop in. Carve. Fly.Recorded Studio result
Actual DeepAny Studio start-and-end-frame result: a skateboarder drops into a bowl, carves the wall and flips, with a camera that swoops with the rider.

Video / Studio result

Drop in. Carve. Fly..

Actual DeepAny Studio start-and-end-frame result: a skateboarder drops into a bowl, carves the wall and flips, with a camera that swoops with the rider.

Example direction

Frame alignment: Picture 1 is the opening frame at zero seconds and Picture 2 is the final frame at five seconds. integrated_multimodal_description: [Shot 1] A high-energy skate shot: the skateboarder pushes off the top lip and drops into the concrete bowl, carves hard up the opposite wall with a spray of grit, launches high into the air and pops a kickflip as the board spins beneath the feet. The camera swoops low and fast alongside the rider with strong motion blur on the walls, then rises to catch the peak of the jump against the golden sun. Continuous powerful movement, flying hair and clothing. No text, logos or watermark. overall_soundscape: Skateboard wheels roaring on concrete, a sharp pop of the board and a landing clack. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Start and end frames
Size
1080 × 1080 · 1:1
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed
Startframe input
Recorded startFrame input for Drop in. Carve. Fly. Download startFrame input
Lastframe input
Recorded lastFrame input for Drop in. Carve. Fly. Download lastFrame input

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / Outrunning the stormRecorded Studio result
Actual DeepAny Studio Text to Video result: a five-second tracking shot of an astronaut racing a dust storm across Martian dunes.

Video / Studio result

Outrunning the storm.

Actual DeepAny Studio Text to Video result: a five-second tracking shot of an astronaut racing a dust storm across Martian dunes.

Example direction

integrated_multimodal_description: [Shot 1] An astronaut in a white suit sprints across rust-red Martian dunes as a colossal orange dust storm wall rolls toward them, kicking up clouds of red sand with each stride. The camera tracks low alongside and then swings to a wide shot as the storm engulfs the horizon, wind whipping fabric straps and dust streaming past the visor. Powerful continuous motion, strong parallax, dramatic light. No text, logos or watermark. overall_soundscape: Howling wind, gritty sand hiss, heavy running breaths in a helmet. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Video
Size
1920 × 1080 · 16:9
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / Rain, neon and a hard cornerRecorded Studio result
Actual DeepAny Studio Text to Video result: a five-second 21:9 drift through a rain-soaked neon street, with a camera that whips from front to behind.

Video / Studio result

Rain, neon and a hard corner.

Actual DeepAny Studio Text to Video result: a five-second 21:9 drift through a rain-soaked neon street, with a camera that whips from front to behind.

Example direction

integrated_multimodal_description: [Shot 1] A low-slung sports car drifts around a tight corner on a rain-soaked neon-lit city street at night. The tires throw sheets of spray, reflections of pink and blue neon streak across the wet asphalt and glass, the rear slides wide and then snaps straight as the car accelerates away. The camera whips from a low front three-quarter angle into a fast tracking move behind the car, rain streaking through the headlights. Strong motion blur and continuous speed. No text, logos or watermark. overall_soundscape: Screaming tires, engine revving, rain and spray. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Video
Size
2520 × 1080 · 21:9
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / Fire in the kitchenRecorded Studio result
Actual DeepAny Studio Text to Video result: a five-second kitchen shot with a flaming wok toss and a camera that swings around the chef.

Video / Studio result

Fire in the kitchen.

Actual DeepAny Studio Text to Video result: a five-second kitchen shot with a flaming wok toss and a camera that swings around the chef.

Example direction

integrated_multimodal_description: [Shot 1] A busy professional restaurant kitchen, close and kinetic: a chef in a white jacket tosses a wok, sending a huge ball of fire and glowing vegetables flipping through the air, flames roaring up around the wok, sparks and steam bursting, then catches everything in a single sweep. The camera pushes in fast and swings around the chef in a smooth arc, with blurred cooks and glowing orange heat lamps moving behind. Intense continuous motion and warm firelight. No text, logos or watermark. overall_soundscape: Roaring flames, sizzling wok, clattering pans, kitchen chatter. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Video
Size
1920 × 1080 · 16:9
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / Fast through the pinesRecorded Studio result
Actual DeepAny Studio Text to Video result: a five-second 21:9 low tracking shot of a wolf running through a snowy forest.

Video / Studio result

Fast through the pines.

Actual DeepAny Studio Text to Video result: a five-second 21:9 low tracking shot of a wolf running through a snowy forest.

Example direction

integrated_multimodal_description: [Shot 1] A grey wolf sprints through a snowy pine forest at dusk, snow exploding from its paws with every stride. The camera tracks low and fast alongside the wolf at eye level, weaving between tree trunks with strong parallax, breath steaming from its muzzle, snowflakes whipping through the blue evening light as the wolf leaps a fallen log and keeps running. Continuous powerful motion for the whole shot. No text, logos or watermark. overall_soundscape: Fast padding paws in deep snow, cold wind through pines, heavy breathing. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Video
Size
2520 × 1080 · 21:9
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / Swimming with giantsRecorded Studio result
Actual DeepAny Studio Text to Video result: a five-second underwater tracking shot of a diver beside a whale shark.

Video / Studio result

Swimming with giants.

Actual DeepAny Studio Text to Video result: a five-second underwater tracking shot of a diver beside a whale shark.

Example direction

integrated_multimodal_description: [Shot 1] An underwater cinematic tracking shot: a scuba diver in a black wetsuit swims alongside an enormous whale shark in clear tropical blue water, the camera gliding with them as the giant spotted body slides past with powerful tail strokes. Shafts of sunlight ripple across the shark and the diver, a school of small silver fish scatters and re-forms, bubbles stream upward, and the camera slowly arcs under the shark to reveal its full length. Continuous graceful motion and strong parallax. No text, logos or watermark. overall_soundscape: Muffled underwater ambience, steady breathing bubbles and a soft low whale-shark hum. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Video
Size
1920 × 1080 · 16:9
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio
Video Generator / Lift-offRecorded Studio result
Actual DeepAny Studio Text to Video result: a five-second 21:9 fantasy launch with a sweeping crane camera.

Video / Studio result

Lift-off.

Actual DeepAny Studio Text to Video result: a five-second 21:9 fantasy launch with a sweeping crane camera.

Example direction

integrated_multimodal_description: [Shot 1] An enormous dark-scaled dragon crouches on a jagged cliff at sunrise, then explodes upward with a thunderous wingbeat, rock and snow bursting from the ledge as the huge wings unfurl. The camera cranes up and back in a sweeping move to follow the dragon as it climbs through torn clouds, golden light glowing through the wing membranes, and banks away over a sea of mist. Epic scale and continuous dramatic motion. No text, logos or watermark. overall_soundscape: A deep roar, enormous wing beats, rushing wind. No speech. non_diegetic_music: N/A

Recorded prompt.

Recipe, inputs & settings
Workflow
Text to Video
Size
2520 × 1080 · 21:9
Duration
5 seconds · 24 fps
Video and upscale enhancement
On · 50% each
Banner
Removed

Text only. No reference files are needed.

Loads into a new project without changing your existing drafts. Review before generating. Where supported, a fixed seed helps repeat a result; service updates and hardware can still change the exact output.

Load recipe in Studio