7
Text-to-video options
Start from a written scene, campaign idea, visual brief, or presentation concept when you do not need an existing image to anchor the first frame.
Video model guide
Video Models in ShiMuv — 19 Ways to Create, Animate, Present, or Translate Video
Shi-Studio currently exposes 19 selectable video model options across nine model families. Some begin with text, some begin with your image, and some transform an existing video. You can choose the engine yourself—or simply describe the outcome you want and let Creation Companion help route the work.
Model availability, durations, and supported settings can change as ShiMuv updates production options. This page reflects the current selectable Shi-Studio catalog.
10
Image-to-video options
Use a creator-owned image, performer reference, product visual, artwork, or finished frame as the visual starting point for motion.
2
Video-to-video options
Transform an existing finished video for another audience, including translation and synchronized dubbed delivery.
Choose by creative job
The model is a production choice—not the creative idea itself.
A still portrait that needs subtle motion is a different job from a cinematic text-to-video scene, a talking presenter, or a translated existing video. ShiMuv exposes multiple engines so the production path can match the creator’s goal instead of forcing every request through one model.
Current model catalog
19 selectable video options, explained in plain language.
Text to videoFast
LTX 2.0
Lightricks
Fast text-to-video creation with audio when you want to move quickly from a written scene to motion.
- Duration:
- 6, 8, or 10 seconds
- Framing:
- 16:9, 9:16, 1:1
Text to video
Wan 2.6
Wan AI
Multi-shot text-to-video with audio for scene ideas that benefit from more than one visual beat.
- Duration:
- 5, 10, or 15 seconds
- Framing:
- 16:9, 9:16, 1:1
Image to video
Wan 2.6 (I2V)
Wan AI
Animate a source image while keeping the image as the visual starting point.
- Duration:
- 5, 10, or 15 seconds
- Framing:
- 16:9, 9:16, 1:1
Text to videoPopular
Seedance Pro
ByteDance
Smooth text-to-video motion with audio for directed scenes and polished visual movement.
- Duration:
- 4, 8, or 12 seconds
- Framing:
- 16:9, 9:16, 1:1
Image to video
Seedance Pro (I2V)
ByteDance
Animate a creator-owned image with a short, directed motion brief.
- Duration:
- 4 or 8 seconds
- Framing:
- 16:9, 9:16, 1:1
Image to videoPremium
Seedance 2.0 (I2V)
ByteDance
Cinematic image-to-video with native audio, camera direction, and start/end-frame control.
- Duration:
- 4–15 seconds or automatic
- Framing:
- 16:9, 9:16, 1:1, auto
Image to videoFast
Seedance 2.0 Fast (I2V)
ByteDance
A faster Seedance 2.0 image-to-video path with native audio and end-frame control.
- Duration:
- 4–15 seconds or automatic
- Framing:
- 16:9, 9:16, 1:1, auto
Image to video
PixVerse v4.5
PixVerse
High-quality image-to-video for short motion from a still visual source.
- Duration:
- Up to 5 seconds
- Framing:
- 16:9, 9:16, 1:1
Image to video
Hailuo (MiniMax)
MiniMax
Image-to-video with an emphasis on expressive motion and facial movement.
- Duration:
- Up to 6 seconds
- Framing:
- 16:9, 9:16, 1:1
Text to videoPremium
Kling 2.5 Turbo Pro
Kuaishou
Cinematic text-to-video for shots where visual polish and directed motion are the priority.
- Duration:
- 5 or 10 seconds
- Framing:
- 16:9, 9:16, 1:1
Image to videoPremium
Kling 2.5 Pro (I2V)
Kuaishou
Animate an existing image with cinematic motion and controlled framing.
- Duration:
- 5 or 10 seconds
- Framing:
- 16:9, 9:16, 1:1
Text to videoPremium
Veo 3
Text-to-video with synchronized audio for short scenes that need both picture and sound.
- Duration:
- Up to 8 seconds
- Framing:
- 16:9, 9:16
Image to videoPremium
Veo 3 (I2V)
Animate a still image with the Veo 3 video workflow.
- Duration:
- Up to 8 seconds
- Framing:
- 16:9, 9:16
Text to videoUltra
Sora 2 Pro
OpenAI
Premium cinematic text-to-video for detailed visual direction and short-form scenes.
- Duration:
- 4, 8, or 12 seconds
- Framing:
- 16:9, 9:16, 1:1
Image to videoUltra
Sora 2 Pro (I2V)
OpenAI
Premium image-to-video when a supplied visual should anchor the generated scene.
- Duration:
- 4, 8, or 12 seconds
- Framing:
- 16:9, 9:16, 1:1
Text to video
HeyGen Video Agent
HeyGen
One-prompt agentic video creation that can assemble script, scenes, and an avatar-led presentation.
- Duration:
- Up to 60 seconds
- Framing:
- 16:9, 9:16, 1:1, auto
Image to video
HeyGen Avatar IV (I2V)
HeyGen
Turn a photo into a talking avatar with voice, lip sync, and expressive presentation.
- Duration:
- Up to 60 seconds
- Framing:
- 16:9, 9:16, 1:1
Video to video
HeyGen Translate
HeyGen
Translate and dub an existing video while preserving synchronized spoken delivery across supported languages.
- Duration:
- Up to 5 minutes
- Framing:
- 16:9, 9:16, 1:1
Video to videoFast
HeyGen Translate (Speed)
HeyGen
A faster video-translation path for creators who need multilingual delivery with shorter turnaround.
- Duration:
- Up to 3 minutes
- Framing:
- 16:9, 9:16, 1:1
One source, more than one output
Use the model to make the asset—then keep building from the result.
The example below comes from the ShiMuv WordPress tutorial library. It is shown as a real finished creator workflow example, not as a claim that one specific model produced every frame. The same finished idea can be adapted for widescreen and vertical destinations.
Try this in Creation Companion
Describe the creative job first. Let the engine choice support it.
Attach the source material that matters and tell ShiMuv what you are trying to make. You can ask for a recommendation or explicitly ask to compare the current selectable model options before generation.
- Attach the exact image, video, song, performer reference, product visual, or other creator-owned source.
- State what must remain exact: identity, product label, room layout, wardrobe, audio, text, or brand details.
- Tell ShiMuv where the finished asset will be used so duration and framing match the destination.
Detailed prompt
I want to make a video and I want ShiMuv to help me choose the production path based on the creative job—not just pick a model name at random. SOURCE MATERIAL - Main source: [attach image, video, song, product visual, performer reference, or other authorized media] - What must remain exact: [identity, face, wardrobe, product label, logo, room/set layout, text, source audio, etc.] - Intended destination: [website, YouTube, TikTok, Instagram, ad, presentation, film, music video, etc.] - Preferred framing: [16:9, 9:16, 1:1, or recommend] - Desired duration: [describe] CREATIVE DIRECTION - What should happen: [describe the action] - Camera behavior: [push in, track, orbit, static, handheld, etc.] - Mood / lighting / pace: [describe] - Audio needs: [native audio, song, voiceover, dialogue, silent, translated speech, etc.] BEFORE GENERATION 1. Identify whether this is text-to-video, image-to-video, or video-to-video. 2. If more than one current ShiMuv model option fits, explain the practical tradeoff between the best two or three options for this specific request. 3. Preserve my source material and locked constraints. Do not silently redesign the person, product, set, or audio I said must remain exact. 4. Show the planned duration, framing, source files, and credit-bearing step before I confirm. 5. If the request is longer than one short clip, route it into the appropriate Film or multi-scene workflow instead of silently shortening it. Do not generate or spend credits until I confirm the plan.
What’s next?
Protect character identity
Learn how references, wardrobe anchors, voice, props, and set locks help recurring people and environments stay coherent.
ContinuePlan the video
See how ShiMuv turns images, songs, products, and ideas into directed video workflows.
ContinueEdit the result
Trim, caption, sequence, and prepare alternate cuts after the master video is approved.
ContinueBuild a longer story
Move multi-scene ideas into Film Studio when the project needs characters, dialogue, continuity, and longer runtime.
ContinueFrequently asked questions
How many video model options does ShiMuv currently expose?
ShiMuv currently exposes 19 selectable video model options in Shi-Studio across text-to-video, image-to-video, and video-to-video workflows. Some options are different modes or speed tiers of the same underlying model family.
Do I have to know which model to choose?
No. You can describe the result you want in Creation Companion and let ShiMuv help route the request. The model guide is useful when you want to make the engine choice yourself in Shi-Studio.
Which model should I use for an image I want to animate?
Choose an image-to-video option when the source image must remain the visual anchor. The right choice depends on duration, framing, audio needs, motion style, and how strongly identity or product details need to remain recognizable.
Does a more expensive model automatically make a better video?
Not necessarily. The best result depends on the task. A short product motion shot, a talking avatar, a cinematic scene, and a translated existing video are different jobs. ShiMuv separates those workflows so you can choose based on the creative outcome rather than price alone.