Models
Video models
Passports of Clipia video models — Veo, Kling, Seedance, Wan, HappyHorse, Hailuo, Grok and more: modes, duration, quality, sound, prices and picks by task.
Clipia runs video models from six developers — Google, Kuaishou, ByteDance, Alibaba, MiniMax and xAI — plus Clipia's own upscaler. Below is which model to pick for a task and a passport for each one: modes, duration, quality, sound, strengths and price in credits. Step-by-step instructions for each mode are on the Video Studio page.
Prices
Prices on this page come from the service pricing and are refreshed together with the documentation. The exact cost is always shown on the run button before any credits are charged.
Which model to choose
| Task | Start with | For better or different results |
|---|---|---|
| A first clip, just trying it out | Grok Imagine Video 1.5 — 5 credits for 5 s at 480p | Seedance 1.5 Pro — 3 credits for 4 s at 480p |
| A text-to-video clip with sound | Veo 3.1 Lite — 11 credits per clip, sound included | Seedance 2.0 Mini, MiniMax H3, Wan 3.0 |
| Top quality | Veo 3.1 Quality | Kling 3, Kling O3, Seedance 2.5, Seedance 2.0 Pro, Wan 3.0 |
| A video from your photo | Grok Imagine Video 1.5 | Hailuo 2.3 Fast, Kling 3 |
| A transition from one frame to another | Veo 3.1 Lite | Kling 2.5 Turbo, Veo 3.1 Fast, Wan 2.7, MiniMax H3, Wan 3.0 |
| A character or product from several images | Veo 3.1 Fast | HappyHorse R2V, Gemini Omni, Seedance 2.0, Kling O3 |
| A long clip, 15 s+ | Seedance 2.5 or Wan 3.0 — up to 30 s | Kling 3, Kling O3, Seedance 2.0 — up to 15 s |
| A multi-scene story with one character | Kling 3 — custom scenes and characters | Turnkey video |
| Animate an old photo | Grok Imagine Video 1.5 | Hailuo 2.3 Fast, Seedance 1.5 Pro |
| A talking photo | Kling Lip Sync | — |
| Change a finished clip | HappyHorse Edit | Kling O3 Transform, Gemini Omni |
| Repeat the motion and camera of a reference | Seedance 2.0 Mini | Seedance 2.0 Pro, Kling O3 R2V, Seedance 2.5, Wan 3.0 R2V |
| Transfer a dance or motion onto your character | Kling 3.0 Motion Control | Kling 2.6 Motion Control — cheaper |
| A clip from a document or presentation | Wan 3.0 | — |
| Sharpen a finished clip | Clipia 4K Upscale | — |
All models at a glance
| Model | Developer | Length | Quality | Sound | Modes |
|---|---|---|---|---|---|
| Grok Imagine Video 1.5 | xAI | 5–15 s | 480p, 720p | no | text, photo, animation |
| Veo 3.1 Lite | 4–8 s | 720p | included | text, photo, first and last frame | |
| Veo 3.1 Fast | 4–8 s | 720p, 1080p, 4K | included | text, photo, frames, references | |
| Veo 3.1 Quality | 4–8 s | 720p, 1080p, 4K | included | text, photo | |
| Gemini Omni | 4–10 s | 720p, 1080p, 4K | yes | text, photo, references, editing | |
| Kling 3 | Kuaishou | 3–15 s | 720p, 1080p, 4K | extra charge | text, photo, scenes, characters |
| Kling 2.6 | Kuaishou | 5–10 s | 720p | extra charge | text, photo, animation |
| Kling 2.5 Turbo | Kuaishou | 5–10 s | — | — | first and last frame |
| Kling O3 | Kuaishou | 3–15 s | 720p, 1080p, 4K | extra charge, included in 4K | text, photo, references |
| Kling O3 Transform | Kuaishou | matches source | 720p, 1080p, 4K | — | editing with a prompt |
| Kling O3 R2V | Kuaishou | matches reference, 3–15 s | — | — | from a reference |
| Kling Lip Sync | Kuaishou | 3–15 s, matches recording | 720p | speech from recording | talking photo |
| Kling 3.0 Motion Control | Kuaishou | 3–30 s, matches video | 720p, 1080p | — | motion control |
| Kling 2.6 Motion Control | Kuaishou | 3–30 s, matches video | 720p, 1080p | — | motion control |
| Seedance 1.5 Pro | ByteDance | 4–12 s | 480p, 720p, 1080p | extra charge | text, photo, animation |
| Seedance 2.0 Mini | ByteDance | 5–15 s | 480p, 720p | included | text, references, from a reference |
| Seedance 2.0 Fast | ByteDance | 5–15 s | 480p, 720p | included | text, references, from a reference |
| Seedance 2.0 Pro | ByteDance | 5–15 s | 480p, 720p, 1080p | included | text, references, from a reference |
| Seedance 2.5 | ByteDance | 4–30 s | 480p, 720p, 1080p | included | references, from a reference |
| Wan 2.7 | Alibaba | 5–15 s | 720p, 1080p | track from a file | text, photo, frames, opening clip |
| Wan 3.0 | Alibaba | 2–30 s, Auto | 480p, 720p, 1080p | yes | text, photo, frames, document |
| Wan 3.0 R2V | Alibaba | 2–30 s | 480p, 720p, 1080p | yes | from a reference |
| HappyHorse 1.0 | Alibaba | 5–15 s | 720p, 1080p | no | text, photo |
| HappyHorse R2V | Alibaba | 5–15 s | 480p–1080p | no | references |
| HappyHorse Edit | Alibaba | matches source, 2–15 s | 480p–1080p | no | editing with a prompt |
| MiniMax H3 | MiniMax | 5–15 s | 480p, 768p | included | text, photo, frames |
| Hailuo 2.3 Fast | MiniMax | 6–10 s | 768p, 1080p | — | photo only |
| Hailuo 2.3 | MiniMax | 6–10 s | 768p, 1080p | — | photo only |
| Clipia 4K Upscale | Clipia | 1–60 s | 2×, 4× | kept | upscaling |
"Length" is the range of durations you can choose; the form shows the exact values for each model. If a mode is not listed, the model does not support it.
How the price is calculated
For most models the price is a cell of a "duration × quality" table, plus a sound surcharge if sound is paid and you turned it on. The exceptions:
- Veo 3.1 — one price per clip, depending only on quality.
- Gemini Omni — a per-clip price by duration; editing is one price per clip.
- Motion control — for every second of the motion video.
- Kling Lip Sync — by the length of the voice recording.
- Editing with a prompt — by source seconds; HappyHorse Edit and Kling O3 Transform have a minimum price.
- From a reference with Seedance and Wan 3.0 R2V — seconds of the new video plus seconds of the reference.
- Upscaling — by clip length and scale factor.
The price on the button is a server quote valid for 10 minutes. The "from N" label in the model list is a guide. How credits are reserved and refunded is covered in Credits and plans.
Veo 3.1 Lite
- Modes: text to video and image to video, first and last frame.
- Length and quality: 4–8 seconds, 720p.
- Sound: on and included in the price.
- Strengths: the most affordable way to get a clip with sound; one price for a clip of any length; the default model for transitions between two frames.
- Price: 11 credits per clip.
Veo 3.1 Fast
- Modes: text to video and image to video, frames, references — up to 3 images, an 8-second clip in 16:9 or 9:16.
- Length and quality: 4–8 seconds, 720p, 1080p or 4K.
- Sound: on and included in the price.
- Strengths: the fast Veo with reference support; the default model for the References mode.
- Price: per clip — 22 credits at 720p, 24 credits at 1080p, 38 credits at 4K.
Veo 3.1 Quality
- Modes: text to video and image to video.
- Length and quality: 4–8 seconds, 720p, 1080p or 4K.
- Sound: on and included in the price.
- Strengths: the best quality among Veo models, part of the Best quality selection.
- Price: per clip — 92 credits at 720p, 94 credits at 1080p, 118 credits at 4K.
Gemini Omni
- Modes: text to video and image to video, references, editing with a prompt — up to 5 images; the source for editing can be up to 30 seconds, with 10 used.
- Length and quality: 4–10 seconds, 720p, 1080p or 4K.
- Sound: yes.
- Strengths: one model for generation, references and editing.
- Price: at 720p — 4 s 30 credits, 10 s 60 credits; at 4K — 4 s 70 credits. Editing a clip at 720p — 80 credits.
Kuaishou (Kling)
Kling 3
- Modes: text to video and image to video; Scene breakdown (up to 6 custom scenes, 3–15 seconds in total); Characters (up to 3, with 2–4 images each).
- Length and quality: 3–15 seconds, 720p, 1080p or 4K.
- Sound: optional, at extra cost: 5 s at 720p with sound — 54 credits.
- Strengths: multi-scene stories with the same character; part of the Best quality selection.
- Price without sound: 720p — 3 s 22 credits, 5 s 36 credits, 10 s 72 credits, 15 s 131 credits; 1080p — 5 s 49 credits, 10 s 87 credits. Other combinations are shown on the button.
Kling 2.6
- Modes: text to video and image to video, photo animation.
- Length and quality: 5 or 10 seconds, 720p.
- Sound: optional, at extra cost: 5 s with sound — 40 credits.
- Price without sound: 5 s — 20 credits, 10 s — 40 credits.
Kling 2.5 Turbo
- Modes: first and last frame.
- Length: 5 or 10 seconds.
- Setting: Prompt adherence — loose, balanced or strict.
- Strengths: an inexpensive smooth transition between two frames.
- Price: 5 s — 16 credits, 10 s — 30 credits.
Kling O3
- Modes: text to video and image to video, references.
- Length and quality: 3–15 seconds, 720p, 1080p or 4K.
- Sound: optional, at extra cost: 5 s at 720p with sound — 65 credits.
- Strengths: part of the Best quality and For references selections.
- Price without sound: 5 s — 50 credits at 720p, 65 credits at 1080p, 240 credits at 4K; the price grows in proportion to length.
Kling O3 Transform
- Mode: editing with a prompt — up to 4 style images.
- Length and quality: matches the source; 720p, 1080p or 4K.
- Price: by source seconds; 5 s — 75 credits at 720p, 100 credits at 1080p, 240 credits at 4K. A short source costs at least 45 credits.
Kling O3 R2V
- Mode: from a reference — one 3–15 second reference clip (MP4 or MOV, up to 100 MB, side 700–4553 px, 24–60 fps) and at least one image.
- Length: matches the reference.
- Price: shown on the button, depends on the reference length.
Kling Lip Sync
- Mode: talking photo — a portrait and a voice recording in MP3, WAV, OGG or AAC, up to 15 seconds.
- Length and quality: matches the recording, 3–15 seconds; 720p.
- Price: 3 s — 30 credits, 5 s — 40 credits, 10 s — 80 credits, 15 s — 150 credits.
Kling 3.0 Motion Control
- Mode: motion control — transfers a person's motion from a video onto the character in a photo.
- Length and quality: matches the video, 3–30 seconds; 720p or 1080p.
- Strengths: the default model for motion transfer.
- Price: grows with the video length; at 720p 3 s — 18 credits, 10 s — 60 credits; at 1080p 10 s — 110 credits.
Kling 2.6 Motion Control
- Mode: motion control.
- Length and quality: 3–30 seconds, 720p or 1080p.
- Strengths: half the price of Kling 3.0 Motion Control at 720p.
- Price: grows with the video length; at 720p 3 s — 9 credits, 10 s — 30 credits; at 1080p 10 s — 51 credits.
ByteDance (Seedance)
Seedance 1.5 Pro
- Modes: text to video and image to video, photo animation; Fixed camera setting.
- Length and quality: 4–12 seconds, 480p, 720p or 1080p.
- Sound: optional, at extra cost: 4 s at 480p with sound — 6 credits.
- Strengths: the lowest entry price — from 3 credits.
- Price without sound: 480p — 4 s 3 credits, 12 s 9 credits; 720p — 4 s 6 credits, 12 s 18 credits. The 1080p price is shown on the button.
Seedance 2.0 Mini
- Modes: text to video with Web search, references, from a reference (up to 3 clips, up to 9 images, audio references).
- Length and quality: 5–15 seconds, 480p or 720p.
- Sound: on and included in the price.
- Strengths: the default model for From a reference; the most affordable Seedance 2.0.
- Price: 480p — 5 s 17 credits, 15 s 51 credits; 720p — 5 s 37 credits, 15 s 110 credits.
Seedance 2.0 Fast
- Modes and limits: same as Seedance 2.0 Mini.
- Length and quality: 5–15 seconds, 480p or 720p.
- Sound: on and included in the price.
- Strengths: the model Turnkey video uses for its scenes.
- Price: 480p — 5 s 28 credits, 15 s 84 credits; 720p — 5 s 47 credits, 15 s 140 credits.
Seedance 2.0 Pro
- Modes and limits: same as Seedance 2.0 Mini.
- Length and quality: 5–15 seconds, up to 1080p.
- Sound: on and included in the price.
- Strengths: the best quality among Seedance 2.0 models, part of the Best quality selection.
- Price: 480p — 5 s 34 credits, 15 s 102 credits; 720p — 5 s 57 credits, 15 s 171 credits; 1080p — 5 s 143 credits.
Seedance 2.5
- Modes: references, from a reference — up to 10 clips (up to 30 seconds in total), up to 30 images, up to 10 audio references.
- Length and quality: 4–30 seconds, up to 1080p.
- Sound: on and included in the price.
- Strengths: the longest clips and the most references; part of the Best quality selection.
- Price: 480p — 5 s 53 credits, 30 s 180 credits; 720p — 5 s 88 credits, 30 s 397 credits; 1080p — 5 s 142 credits.
Alibaba (Wan and HappyHorse)
Wan 2.7
- Modes: text to video and image to video, first and last frame, opening clip (a 2–10 second MP4 or MOV that the model continues).
- Length and quality: 5–15 seconds, 720p or 1080p.
- Sound: your own track from an MP3 or WAV file up to 15 seconds; it does not change the price.
- Setting: Prompt enhancement, on by default.
- Strengths: the only model that continues your own clip.
- Price: 720p — 5 s 24 credits, 15 s 70 credits; 1080p — 5 s 40 credits, 15 s 114 credits.
Wan 3.0
- Modes: text to video and image to video, first and last frame, a clip from a document (PDF, Word, Excel, PowerPoint and more, up to 100 MB and 50 pages).
- Length and quality: 2–30 seconds or Auto, 480p, 720p or 1080p.
- Sound: yes.
- Setting: Speed — normal or fast.
- Strengths: clips up to 30 seconds, clips from documents; part of the Best quality selection.
- Price: grows with length; 5 s — 16 credits at 480p, 32 credits at 720p, 64 credits at 1080p; the shortest clip, 2 s at 480p, is 7 credits. Auto reserves as for 30 seconds and charges for the actual length.
Wan 3.0 R2V
- Mode: from a reference — up to 5 clips (1–15 seconds, up to 15 in total), up to 10 images, up to 5 audio references; the reference and the new video together up to 30 seconds.
- Quality: 480p, 720p or 1080p.
- Price: same as Wan 3.0 — by seconds of the new video and the reference.
HappyHorse 1.0
- Modes: text to video and image to video.
- Length and quality: 5–15 seconds, 720p or 1080p.
- Sound: no.
- Price: depends on length and resolution; the exact amount is on the run button.
HappyHorse R2V
- Mode: references — up to 9 images.
- Length and quality: 5–15 seconds, 480p–1080p.
- Sound: no.
- Strengths: recommended for the References mode.
- Price: depends on length and resolution; the exact amount is on the run button.
HappyHorse Edit
- Mode: editing with a prompt — a 2–15 second source, up to 9 style images.
- Quality: 480p–1080p.
- Sound: no.
- Strengths: the default model for editing.
- Price: by source seconds, using the same table as HappyHorse R2V; there is a minimum price.
MiniMax (Hailuo)
MiniMax H3
- Modes: text to video and image to video, first and last frame.
- Length and quality: 5–15 seconds, 480p or 768p.
- Sound: on and included in the price.
- Price: 480p — 5 s 16 credits, 15 s 48 credits; 768p — 5 s 32 credits, 15 s 96 credits.
Hailuo 2.3 Fast
- Modes: photo only, photo animation.
- Length and quality: 6 or 10 seconds, 768p or 1080p.
- Strengths: inexpensive, fast image to video; recommended for photo animation.
- Price: 6 s at 768p — 11 credits, 10 s at 768p — 19 credits, 6 s at 1080p — 19 credits.
Hailuo 2.3
- Modes: photo only.
- Length and quality: 6 or 10 seconds, 768p or 1080p.
- Price: 6 s at 768p — 18 credits, 10 s at 768p — 33 credits, 6 s at 1080p — 29 credits.
xAI
Grok Imagine Video 1.5
- Modes: text to video and image to video, photo animation.
- Length and quality: 5–15 seconds, 480p or 720p.
- Sound: no.
- Strengths: the default model in the Video Studio; the 5 welcome credits cover a 5-second 480p clip.
- Price: 480p — 5 s 5 credits; 720p — 5 s 9 credits, 10 s 18 credits, 15 s 27 credits.
Clipia
Clipia 4K Upscale
- Mode: upscaling a finished clip.
- Length: clips of 1–60 seconds.
- Scale: 2× — up to 1080p–1440p; 4× — up to 4K, for sources of 720p or lower.
- Sound: kept.
- Price: by clip length; 2× — 5 s 10 credits, 60 s 120 credits; 4× — 5 s 15 credits, 60 s 180 credits.
Where the models are available
- Video Studio (
/create-video) — every model and mode on this page. - Model pages at
/create-video/<model>— a description, examples and a shortcut to generating with that model. - Public API — models for developers; the list and prices in the API may differ from the app.
Every model is available on every plan. For a comparison with image and audio models, see How to choose a model.
Image models
All image models in Clipia — developers, strengths, aspect ratios and resolution, how many photos they accept, modes and price in credits. How to choose one.
Audio and speech models
The Clipia models that voice text and make music — developers, languages, voices, prices in credits and tips on choosing a model and a voice.