Image to video models, and what each one costs
Image to video is the mode that matters when something already exists and has to survive: a product, a face, a room, a piece of artwork. The model is not inventing the subject, it is moving the one you gave it, and the prompt directs the movement rather than describing the scene.
67 models on Musevate animate an uploaded image. The cheapest costs 5 credits for five seconds, 32 of them generate sound, and 60 also work from a prompt alone, so you are not choosing a different model when you switch modes.
67 models, read from the live registry when this page loaded.
| # | Model | 5 seconds | Resolution | Longest | Sound | Example |
|---|---|---|---|---|---|---|
| 1 | P-Video Fast iteration, with a draft mode. | 5 cr | 1080p, 720p | 20s | No. It takes an audio track and animates to it | Watch |
| 2 | Seedance 1.5 Pro Joint audio and video, following complex instruction. | 8 cr | 1080p, 720p, 480p | 12s | On request, with audio switched on | Watch |
| 3 | Grok Imagine 1.5 Lite Grok Imagine 1.5 for a fraction of the price. | 10 cr | 1080p, 720p, 480p | 15s | Always, as part of the generation | Watch |
| 4 | P-Video-2 Pruna's quality-focused P-Video, up to 1080p. | 10 cr | 1080p, 720p | 20s | Always, as part of the generation | Watch |
| 5 | Hailuo 02 Hailuo 2, strong on physical realism. | 12 cr | 1080p, 768p | 10s | No | Watch |
| 6 | LTX 2.5 Fast Built for speed, with native audio and high resolutions. | 12 cr | 4k, 1440p, 1080p, 720p | 20s | On request, with audio switched on | Watch |
| 7 | P-Video-2 Pro Pruna's top model, run in its quality mode. | 13 cr | 4k, 1440p, 1080p, 768p, 480p | 15s | Always, as part of the generation | Watch |
| 8 | MiniMax H3 Max H3 retrained for closer prompt adherence. | 14 cr | 1080p, 768p, 480p | 10s | Always, as part of the generation | Watch |
| 9 | Grok Imagine Video xAI's Grok Imagine video model. | 15 cr | 720p, 480p | 15s | Always, as part of the generation | Watch |
| 10 | Hailuo 2.3 High fidelity model tuned for realistic human motion. | 15 cr | 1080p, 768p | 10s | No | Watch |
| 11 | Kandinsky 6.0 Lite Kandinsky's fast model: five seconds, with sound. | 17 cr | 480p | 5s | On request, with audio switched on | Watch |
| 12 | Kling v2.6 Kling 2.6 Pro, cinematic image-to-video with fluid motion. | 18 cr | 1080p | 10s | On request, with audio switched on | Watch |
| 13 | Kling v2.5 Kling 2.5 Turbo Pro, professional output at speed. | 20 cr | 1080p | 10s | No | Watch |
| 14 | Pixelcut Looping Video A product photo, turned into a seamless loop. | 20 cr | 1080p, 768p, 480p | 15s | No | Watch |
| 15 | Vidu Q4 Image-led video with sound, up to 4K. | 20 cr | 4k, 1440p, 1080p, 720p, 540p | 16s | Always, as part of the generation | Watch |
| 16 | Wan 2.2 The dependable all-rounder for everyday production work. | 20 cr | 720p, 580p, 480p | 10s | No | Watch |
| 17 | Kling v3 Cinematic framing with fluid, believable movement. | 22 cr | 4k, 1080p, 720p | 15s | On request, with audio switched on | Watch |
| 18 | LTX 2.3 Video Fast The previous LTX generation, tuned for speed. | 23 cr | 4k, 1440p, 1080p | 20s | On request, with audio switched on | Watch |
| 19 | Gemini Omni Flash 1.1 Google's Gemini Omni Flash 1.1, with sound. | 25 cr | 4k, 1080p, 720p, 360p | 10s | Always, as part of the generation | Watch |
| 20 | Video 01 The original Hailuo model, in 6 second clips. | 25 cr | 720p | 5s | No | Watch |
| 21 | Video 01 Director Video-01 with explicit camera direction. | 25 cr | 720p | 5s | No | Watch |
| 22 | Wan 2.7 Wan 2.7, with first and last frame control. | 25 cr | 1080p, 720p | 15s | Always, as part of the generation | Watch |
| 23 | Wan 3.0 Smoother motion and cleaner scenes than Wan 2.2, up to 1080p. | 25 cr | 1080p, 720p, 480p | 10s | On request, with audio switched on | Watch |
| 24 | Veo 3 Fast Veo 3 at a lower price and a shorter wait. | 27 cr | 1080p, 720p | 8s | On request, with audio switched on | Watch |
| 25 | Veo 3.1 Fast Veo 3.1 accelerated, for iterating on a shot quickly. | 27 cr | 4k, 1080p, 720p | 8s | On request, with audio switched on | Watch |
| 26 | Gen 4.5 Runway Gen-4.5, strong motion quality and prompt adherence. | 30 cr | 720p | 10s | No | Watch |
| 27 | 30 cr | 1080p, 720p, 540p, 360p | 8s | No | Watch | |
| 28 | 30 cr | 1080p, 720p, 540p, 360p | 8s | No | Watch | |
| 29 | 30 cr | 1080p, 720p, 540p, 360p | 8s | No | Watch | |
| 30 | Seedance 2.0 Mini The smallest Seedance 2.0 tier. | 30 cr | 720p, 480p | 15s | On request, with audio switched on | Watch |
| 31 | Wan 3.0 Prime The flagship Wan tier, for shots that carry the piece. | 35 cr | 1080p, 720p, 480p | 30s | On request, with audio switched on | Watch |
| 32 | Ray 2 720p Luma Ray 2 at 720p, in 5 and 9 second clips. | 45 cr | 720p | 9s | No | Watch |
| 33 | Seedance 2.0 Fast Seedance 2.0 at speed. | 45 cr | 720p, 480p | 15s | On request, with audio switched on | Watch |
| 34 | Veo 3.1 The current flagship, strongest on realism and natural motion. | 50 cr | 4k, 1080p, 720p | 8s | On request, with audio switched on | Watch |
| 35 | Kandinsky 6.0 Pro Kandinsky Lab's flagship, with sound. | 70 cr | 480p | 5s | On request, with audio switched on | Watch |
| 36 | Kling v2.0 Kuaishou's Kling 2.0, in 5 and 10 second clips. | 70 cr | 720p | 10s | No | Watch |
| 37 | Kling v2.1 Master The premium 2.1 tier, with stronger dynamics. | 70 cr | 1080p | 10s | No | Watch |
| 38 | Seedance 2.5 Single shot clips up to 30 seconds, with audio. | 115 cr | 1080p, 720p, 480p | 30s | On request, with audio switched on | Watch |
| 39 | Veo 2 Google Veo 2, faithful to written direction. | 120 cr | 720p | 8s | No | Watch |
| 40 | 5 cr | 480p | 5s | No | Not yet | |
| 41 | Seedance v1 Pro Fast Seedance 1 Pro text-to-video, accelerated. | 6 cr | 1080p, 720p, 480p | 12s | No | Not yet |
| 42 | MiniMax H3 Max Turbo H3 Max with the brakes off, at half the cost. | 8 cr | 1080p, 768p, 480p | 15s | No | Not yet |
| 43 | Ray Flash 2 540p The cheapest Luma tier, at 540p. | 9 cr | 540p | 9s | No | Not yet |
| 44 | Veo 3.1 Lite The lightest Veo 3.1 tier. | 9 cr | 1080p, 720p | 8s | On request, with audio switched on | Not yet |
| 45 | Kling v2.5 Standard The Kling 2.5 look, at the standard tier. | 10 cr | 720p | 10s | No | Not yet |
| 46 | 10 cr | 480p | 10s | No | Not yet | |
| 47 | 10 cr | 720p, 480p | 16s | No | Not yet | |
| 48 | Seedance 1 Lite The light Seedance tier, for text and image driven clips. | 10 cr | 1080p, 720p, 480p | 12s | No | Not yet |
| 49 | Hailuo 2.3 Fast Hailuo 2.3 with lower latency, keeping its motion quality. | 12 cr | 1080p, 768p | 10s | No | Not yet |
| 50 | MiniMax H3 MiniMax H3: video with sound from a prompt, a frame or references. | 14 cr | 4k, 1440p, 768p, 480p | 10s | Always, as part of the generation | Not yet |
| 51 | Kling 1.6 Kling 1.6 Standard, Kuaishou's release of December 2024. | 15 cr | 720p | 10s | No | Not yet |
| 52 | Kling 2.1 Kling 2.1 image-to-video, a proven older generation. | 15 cr | 720p | 10s | No | Not yet |
| 53 | PixVerse v6 Pixverse v6, strong on stylised and animated looks. | 15 cr | 1080p, 720p, 540p, 360p | 15s | No | Not yet |
| 54 | Seedance 1.0 Pro Seedance 1 Pro for text-to-video. | 15 cr | 1080p, 720p, 480p | 12s | No | Not yet |
| 55 | Grok Imagine 1.5 Expressive motion from a still, with sound. | 20 cr | 1080p, 720p, 480p | 15s | Always, as part of the generation | Not yet |
| 56 | Video 01 Live Trained for Live2D and illustrated characters. | 25 cr | 720p | 5s | No | Not yet |
| 57 | Wan 2.5 Preview build of Wan 2.5. | 25 cr | 1080p, 720p, 480p | 10s | No | Not yet |
| 58 | Kling O3 The Kling O3 standard tier. | 30 cr | 4k, 1080p, 720p | 15s | On request, with audio switched on | Not yet |
| 59 | Kling Video v3 Standard Turbo Kling v3 quality at turbo speed. | 30 cr | 720p | 15s | Always, as part of the generation | Not yet |
| 60 | Gemini Omni Flash Google's Gemini Omni Flash, the release before 1.1. | 35 cr | 720p | 10s | No | Not yet |
| 61 | Happy Horse Alibaba's Happy Horse text-to-video model. | 35 cr | 1080p, 720p | 15s | No | Not yet |
| 62 | Happy Horse 1.1 Happy Horse 1.1, refined over the first release. | 35 cr | 1080p, 720p | 15s | No | Not yet |
| 63 | Kling Video v3 Turbo Pro The professional Kling v3 tier, accelerated. | 35 cr | 1080p | 15s | Always, as part of the generation | Not yet |
| 64 | FLUX 3 Video Coherent, physically plausible motion from a single frame. | 45 cr | 4k, 1440p, 1080p, 720p | 20s | On request, with audio switched on | Not yet |
| 65 | Veo 3 Google's Veo 3, with generated audio. | 50 cr | 1080p, 720p | 8s | On request, with audio switched on | Not yet |
| 66 | Seedance 2.0 ByteDance's multimodal generation with native audio. | 55 cr | 4k, 1080p, 720p, 480p | 15s | On request, with audio switched on | Not yet |
| 67 | Wan 2.1 Accelerated Wan 2.1 image-to-video at 720p. | 60 cr | 720p, 480p | 5s | No | Not yet |
Questions
What is image to video?
You upload a still image and the model animates it, using your prompt to direct the movement rather than to invent the subject. The face, the product, the room stay recognisably the ones you supplied. That is the difference from text to video, where everything in the frame is invented and nothing is guaranteed to match anything you already have.
How many image to video models are there on Musevate?
67 of the 73 models animate an uploaded image. 60 of them also work from a prompt alone, so the same model can be used either way, and 32 generate sound as well.
Will the model keep my product or my face looking right?
That is exactly what image to video is for, and it is the main thing that separates these models from one another. Holding a likeness through movement is harder than starting a scene from nothing, and models vary in how far they will drift. Short clips and modest movement drift least, whichever model you choose.
What image should I upload?
The sharpest and largest version you have, in the aspect ratio you want back. A model animating a small or soft image has less to work with and the result shows it. Musevate strips the metadata from anything you upload before it goes anywhere, so location and camera details in the file do not travel with it.
Can I set the last frame as well as the first?
Some models accept an ending image as well as a starting one and generate the movement between them, which is the most control available over what happens in the clip. Each model page states whether that model takes one, because it is a property of the model rather than of the platform.
Try any of them on one account
Musevate routes each generation to the strongest compatible model, and every model above is available without a separate subscription.
Create your first video