图像转视频
上传一张图像，选择视频模型并提交相同格式的请求。Muapi 统一了图生视频模型的鉴权、异步结果和价格，让你可以快速比较动画效果。
- →使用 Kling、Wan、Seedance、Veo、Pixverse 和更多模型让照片动起来
- →通过 URL 或上传文件提供参考图像
- →所有模型均使用提交后轮询和可选 webhook 流程
- →按成功生成计费，无需单独注册模型提供商
快速开始
此类别中的所有模型都使用相同的提交后轮询 API。将 kling-o1-standard-image-to-video 替换为下方列表中的任意模型端点。
# 1. Submit
curl -X POST https://api.muapi.ai/api/v1/kling-o1-standard-image-to-video \
-H "x-api-key: $MUAPI_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt":"slow camera dolly forward, gentle wind on leaves"}'
# → {"request_id":"abc123","status":"processing"}
# 2. Poll until completed
curl https://api.muapi.ai/api/v1/predictions/abc123/result \
-H "x-api-key: $MUAPI_API_KEY"排名前 5 的图像转视频模型
| 模型 | 提供商 | 成本 | 适用场景 |
|---|---|---|---|
| grok-imagine-image-to-video | — | $0.150 | Grok Imagine is xAI’s multimodal image-to-video model, capable of animating still images into cinematic videos from 6 to 30 seconds with synchronized ambient audio. It focuses on realism, fluid motion, and expressive lighting transitions while maintaining high generation speed. |
| seedance-2-i2v | — | $0.750 | SD 2.0 is the latest multimodal video generation model by ByteDance, offering advanced camera control, native audio-video sync, and high-resolution output. |
| seedance-2.5-omni-reference | — | $1.700 | Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview). |
| kling-v3.0-pro-image-to-video | — | $0.720 | Kling 3.0 Pro Image-to-Video animates a single input image into a high-quality, realistic video with smooth camera motion, natural physics, and strong temporal consistency. It excels at real-world scenes, human motion, environmental details, and cinematic movement while preserving the original image’s structure and lighting. |
| seedance-2.5-spicy-omni-reference-480p | — | $0.935 | Seedance 2.5 Omni Reference 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, blending multiple reference images, video clips, and audio into a single guided generation. Faster and more cost-effective than the 720p variant. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output. |
全部 170 个模型
veo3-image-to-video
VEO3 I2V animates static images into expressive video sequences, adding lifelike movement while preserving the original composition.
minimax-hailuo-2.3-pro-i2v
Hailuo 2.3 Pro I2V breathes life into still images with stunning motion synthesis and cinematic camera control. Using deep motion understanding, it predicts realistic subject movement, depth, and environmental motion from a single input frame — delivering smooth, film-grade clips.
kling-v2.1-master-i2v
Kling 2.1 Master’s I2V animates a still image into a coherent video sequence. It interprets motion, environment, and context to create realistic, visually stunning video outputs — ideal for animating portraits, scenes, or concept art.
gemini-omni-flash-1-1-image-to-video
Gemini Omni 1.1 Flash Image to Video — animate a starting keyframe image into a video, with an optional ending keyframe image for controlled first/last-frame transitions, and native synchronized audio.
seedance-2.5-omni-reference
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview).
runway-act-two-i2v
Upload a single character image and a driving video — the model transfers facial expressions and head movements from the video onto your image, bringing it to life. It works with photos, illustrations, or stylized portraits, making them speak, blink, and move naturally. Ideal for avatars, AI presenters, digital actors, and story scenes.
seedance-2-i2v-480p
SD 2.0 480p image-to-video generation. Faster and more cost-effective than the 720p variant, ideal for previews and drafts.
seedance-pro-i2v-fast
Seedance Pro Fast is the high-speed image-to-video generation variant from ByteDance’s Seedance series. With this model you upload a reference image and—using a text prompt—generate short, dynamic video clips (typically 3-12 seconds) featuring smooth motion, cinematic camera moves, prompt-accurate actions, and high visual fidelity. It supports resolutions up to 1080p, multiple aspect ratios (16:9, 9:16, etc.), and rapid turnaround—ideal for social content, product motion, storytelling from a still, and fast prototyping.
seedance-2.5-intl-image-to-video-1080p
Seedance 2.5 Image-to-Video is the early-access preview build of the Seedance 2.5 family, animating a single image into video. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
seedance-2-vip-image-to-video-fast-1080p
SD 2 Image-to-Video VIP 1080p Fast by ByteDance. Faster 1080p animation of a still image with priority routing, 4–15 second duration.
wan3.0-prime-reference-to-video
Wan 3.0 Prime Reference to Video generates a higher-fidelity video guided by a prompt plus up to 10 reference images, 5 reference videos, and 5 reference audios for visual and audio coherence.
seedance-2-i2v
SD 2.0 is the latest multimodal video generation model by ByteDance, offering advanced camera control, native audio-video sync, and high-resolution output.
wan2.7-image-to-video
Alibaba WAN 2.7 converts images into videos with optional audio.
wan2.7-reference-to-video
Alibaba WAN 2.7 Reference-to-Video. Reference characters/props to generate new shots.
openai-sora-2-image-to-video
Sora 2’s I2V lets you bring still images to life by animating them into short video clips with natural motion, audio, and visual effects. While realistic portraits of people aren’t allowed at launch, you can use objects, landscapes, stylized characters or scenes. Use detailed prompts for camera movement, atmosphere, and pacing to get the best results.
seedance-2-image-to-video
SD 2 Image-to-Video (Pro) by ByteDance. Animates a start-frame image into a high-quality video with native audio, 4–15 second duration, and 2K resolution.
gemini-omni-flash-1-1-reference-to-video
Gemini Omni 1.1 Flash Reference to Video — generate video referencing up to 7 images and up to 3 short reference video clips (each up to 3 seconds), addressed by tag directly in the prompt.
seedance-2-vip-omni-reference-fast-1080p
SD 2 Omni Reference VIP 1080p Fast by ByteDance. Faster full HD omni-reference video generation with priority routing. Supports image, video, and audio references.
seedance-2-first-last-frame-fast
SD 2 First & Last Frame (Fast) by ByteDance. Quickly generate video that transitions between reference images at reduced cost. Provide 1 or 2 images.
seedance-2-omni-reference-no-video-fast
SD 2 Omni Reference (Fast) by ByteDance. Quickly generate videos using up to 9 image references and up to 3 audio references at reduced cost. Reference images in your prompt with @image1, @image2, etc. and audio with @audio1, @audio2, etc.
seedance-2-vip-image-to-video-fast
SD 2 Image-to-Video VIP Fast by ByteDance. Faster animation of a start-frame image with priority routing, 4–15 second duration, and 2K resolution.
seedance-pro-i2v
Seedance Pro I2V advanced model animates still images into stunning short videos, preserving intricate visual details and applying smooth motion dynamics, ideal for high-end visuals and cinematic edits.
happy-horse-1-image-to-video-1080p
Happy Horse 1.0 Image to Video — bring still images to life with fluid, expressive animation and fine-grained motion control.
kling-v2.1-standard-i2v
Kling 2.1 Standard (developed by Kuaishou) brings static images to life by generating smooth, realistic video clips from a single frame. It captures subtle motion, background dynamics, and camera movement to produce professional-looking animations — ideal for portraits, digital art, and cinematic illustrations.
kling-v2.1-pro-i2v
Kling 2.1 Pro is the high-end version of Kuaishou’s video generation model, offering enhanced realism, longer motion sequences, and cinematic quality. In I2V mode, it animates static images with fluid environmental effects.
wan2.2-image-to-video
Wan 2.2’s I2V mode brings static visuals to life with vivid, expressive animations. It interprets motion, emotion, and background dynamics from a single image to generate smooth and cinematic short videos.
vidu-v2.0-i2v
Vidu's 2.0 model delivers advanced image-based video generation with enhanced lighting, emotion dynamics, and automatic frame interpolation for polished visual content.
openai-sora-2-pro-image-to-video
Sora 2 Pro I2V brings still images to life, transforming them into short videos with natural motion, realistic lighting, and synchronized audio. Upload your image, describe the movement (camera motion, subject action, ambience), add optional dialogue or sound effects, and watch it animate. Ideal for cinematic reveals, promo videos, social content, or storytelling from a static photo.
minimax-hailuo-02-standard-i2v
Transforms an image into video with light, natural motion. Great for social media, quick animations, and previews.
wan2.1-reference-video
WAN 2.1 is an advanced AI model that transforms one or more reference images into a coherent, animated video. By combining characters, objects, or environments from multiple images, it creates smooth motion sequences while preserving realism, style, and fine details.
kling-v2.5-turbo-pro-i2v
Kling 2.5 Turbo Pro: Top-tier image-to-video generation with unparalleled motion fluidity, cinematic visuals, and exceptional prompt precision.
leonardoai-motion-2.0
Motion 2.0 is Leonardo.AI's cutting-edge model for creating high-quality 5-second videos from text prompts. It offers enhanced control over animation, including camera movements, lighting, and scene dynamics.
grok-imagine-image-to-video
Grok Imagine is xAI’s multimodal image-to-video model, capable of animating still images into cinematic videos from 6 to 30 seconds with synchronized ambient audio. It focuses on realism, fluid motion, and expressive lighting transitions while maintaining high generation speed.
kling-o1-reference-to-video
Kling O1’s Reference-to-Video mode generates a dynamic video using one or multiple reference images as the visual foundation. It preserves identity, style, composition, and key visual details from the references while adding realistic camera motion, environment dynamics, and scene animation.
wan2.6-image-to-video
WAN 2.6 Image-to-Video converts a single still image into a smooth, cinematic video clip. It preserves the original image’s composition, lighting, and style while adding natural motion, depth parallax, atmospheric effects, and gentle camera movement.
ovi-image-to-video
Ovi is a unified audio–video generation model that can transform a static image plus a descriptive prompt into a short video with synchronized audio. It supports both text-to-video and image-conditioned video inputs. With built-in lip sync, background audio / sound effects, and dialogue support, Ovi brings still visuals to life in cinematic fashion. Videos are generated in 540p resolution.
seedance-v1.5-pro-i2v
Seedance v1.5 Pro Image-to-Video converts a single still image into a smooth cinematic video clip. It preserves the original image’s composition, subject identity, and lighting while adding controlled camera motion, natural parallax, and environmental animation. This mode balances visual quality and motion complexity, making it ideal for cinematic scenes, fantasy worlds, sci-fi environments, and storytelling shots.
seedance-2-omni-reference
SD 2.0 Omni Reference — generate videos with visual consistency using reference images, videos, and audio. Maintain character identity, style, and scene continuity. Supports up to 9 images, 3 video clips, and 3 audio clips. Use @image1, @video1, @audio1 syntax in your prompt.
veo3.1-fast-image-to-video
Veo 3.1 Fast is an optimized version of Google’s Veo 3.1 AI that transforms static images into dynamic 8-second videos at higher speed. It preserves visual fidelity while enabling rapid generation, making it ideal for social media clips, storyboards, and quick creative previews.
vidu-q2-reference
Vidu Q2 Reference Video generates breathtaking cinematic clips from text prompts guided by multiple reference images. Each image refines the model’s understanding of subject, environment, and visual tone — ensuring perfect consistency in appearance and motion across every frame.
minimax-hailuo-2.3-standard-i2v
Hailuo 2.3 Standard I2V converts still images into visually immersive motion clips with stable dynamics and realistic movement. It provides a balanced mix of quality, speed, and coherence. In 768p video generation.
minimax-hailuo-2.3-fast
Minimax Hailuo 2.3 Fast is the lightweight, high-speed version of the Hailuo 2.3 family — designed for creators who need instant video generation with cinematic motion and scene consistency. In 768p video generation.
kling-v2.5-turbo-std-i2v
Kling 2.5 Turbo Std: Top-tier image-to-video generation with unparalleled motion fluidity, cinematic visuals, and exceptional prompt precision.
kling-v3.0-standard-image-to-video
Kling 3.0 Standard Image-to-Video animates a single input image into a short, realistic video with smooth, stable motion. It prioritizes temporal consistency, natural physics, and subtle camera movement, making it ideal for everyday scenes, travel moments, people, vehicles, and calm cinematic shots.
kling-v3.0-pro-image-to-video
Kling 3.0 Pro Image-to-Video animates a single input image into a high-quality, realistic video with smooth camera motion, natural physics, and strong temporal consistency. It excels at real-world scenes, human motion, environmental details, and cinematic movement while preserving the original image’s structure and lighting.
veo3.1-lite-image-to-video
Veo 3.1 Lite is a lightweight variant of Google's Veo 3.1 model designed for faster, more accessible video generation from images.
kling-v2.6-pro-i2v
Kling-v2.6-Pro Image-to-Video transforms a single creative image into a short cinematic video. It preserves the original style, lighting, and composition while adding smooth camera motion, atmospheric effects, and dynamic environmental animation.
pixverse-v5.5-i2v
PixVerse v5.5 I2V transforms a single image into a dynamic cinematic video clip. It adds smooth camera motion, atmospheric animation, natural parallax, and environmental effects while preserving the image’s original art style and composition.
hunyuan-image-to-video
Hunyuan I2V takes a static image and generates realistic video animations by interpreting motion and context. It works well for human portraits, objects, or scenes, adding lifelike movement while maintaining the image's integrity.
wan2.2-spicy-image-to-video
Wan2.2-spicy Image-to-Video transforms a single creative image into a short dynamic video with bold motion, stylized effects, high-contrast lighting, and energy-driven animations. The “spicy” variant produces more dramatic movement, more vivid colors, and more expressive visual effects.
seedance-2-omni-reference-no-video
SD 2 Omni Reference by ByteDance. Generate videos using up to 9 image references and up to 3 audio references. Reference images in your prompt with @image1, @image2, etc. and audio with @audio1, @audio2, etc.
kling-o1-standard-image-to-video
Kling O1 Standard Image-to-Video converts a single still image into a short, natural-looking video clip. It preserves the original image’s composition and lighting while adding subtle camera motion, gentle parallax, and light environmental animation. This mode focuses on realism and stability rather than heavy effects, making it ideal for clean cinematic shots, environments, characters, and product visuals.
seedance-v1.5-pro-i2v-fast
Seedance v1.5 Pro Image-to-Video Fast converts a single still image into a short cinematic video with quick generation speed. It preserves the original image’s composition, subject identity, and lighting while adding simple camera motion, light parallax, and subtle environmental animation.
seedance-2.5-intl-image-to-video-4k
Seedance 2.5 Image-to-Video is the early-access preview build of the Seedance 2.5 family, animating a single image into video. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
seedance-2-mini-image-to-video
Seedance 2.0 Mini Image-to-Video is the fastest and most cost-efficient tier in the Seedance 2.0 family, converting still images into smooth cinematic video clips at up to 720p. Roughly 2x faster than Seedance 2.0 Fast, it is purpose-built for high-volume production, rapid iteration, and draft workflows.
ltx-2-19b-image-to-video
LTX-2-19B Image-to-Video animates a single image into a coherent cinematic clip with strong temporal stability. It preserves composition and lighting while adding controlled camera motion, realistic parallax, and subtle environmental dynamics—well suited for grounded scenes, near-future concepts, and story beats.
veo-4-image-to-video
Veo 4 Image to Video — animate any still image with Veo 4's motion synthesis engine, supporting fine-grained camera control and realistic physics at up to 1080p.
seedance-2-vip-image-to-video-1080p
SD 2 Image-to-Video VIP 1080p by ByteDance. Animates a still image into a cinematic 1080p video with priority routing, 4–15 second duration.
kling-v3.0-4k-image-to-video
Kling 3.0 4K Image-to-Video animates a single input image into ultra-high-resolution 3840×2160 video with smooth camera motion, natural physics, and strong temporal consistency. 4K mode delivers the sharpest detail in Kling 3.0 — ideal for cinematic shots, product showcases, and premium content where pixel-level clarity matters.
happy-horse-1-image-to-video-720p
Happy Horse 1.0 Image to Video (720p) — bring still images to life with fluid, expressive animation at 720p output resolution.
seedance-2-vip-first-last-frame-1080p
SD 2 First & Last Frame VIP 1080p by ByteDance. Generate 1080p video that transitions between two reference images with priority routing. Provide 1 image for start-frame-only, or 2 images for both start and end frames.
pixverse-v6-i2v
Animate any image into a video using PixVerse V6. Supports resolutions up to 1080p, durations up to 15 seconds, and prompt-based motion control.
vidu-q2-turbo-image-to-video
Vidu Q2 Turbo Image-to-Video animates a starting image into a fast, prompt-guided clip while preserving subject identity. Built for speed and cost efficiency.
seedance-2.5-first-last-frame
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview).
seedance-2-image-to-video-fast
SD 2 Image-to-Video (Fast) by ByteDance. Quickly animates a start-frame image into video with 4–15 second duration at reduced cost.
seedance-2-first-last-frame
SD 2 First & Last Frame (Pro) by ByteDance. Generate video that transitions between two reference images. Provide 1 image for start-frame-only, or 2 images for both start and end frames.
flux-3-start-end-to-video
FLUX 3 Start-End to Video generates a video transition between a start and end keyframe, with optional native synchronized audio, using Black Forest Labs' unified image/video/audio architecture.
seedance-2-vip-omni-reference-fast
SD 2 Omni Reference VIP Fast by ByteDance. Faster video generation using up to 9 image references, up to 3 video clips, and up to 3 audio references with priority routing. Reference materials in your prompt with @image1…@image9, @video1…@video3, and @audio1…@audio3.
seedance-2-mini-spicy-image-to-video
Seedance 2 Mini Spicy Image-to-Video is the fastest, lowest-cost Spicy-tier image animation, with reduced content-safety filtering on top of Seedance 2 Mini's speed and pricing.
gemini-omni-image-to-video
Gemini Omni Image to Video — animate one or more reference images with a text prompt. Unified reasoning across modalities preserves subject identity and generates synchronized audio natively.
happy-horse-1-reference-to-video-1080p
Happy Horse 1.0 Reference to Video (1080p) - generate expressive 1080p video clips conditioned on 1-9 reference images plus a text prompt.
happy-horse-1-reference-to-video-720p
Happy Horse 1.0 Reference to Video (720p) - generate expressive 720p video clips conditioned on 1-9 reference images plus a text prompt.
grok-imagine-video-1-5-preview
Generate videos from images using the Grok Imagine Video 1.5 Preview model with support for multiple aspect ratios, resolutions, and durations up to 15 seconds.
kling-v3.0-omni-standard-image-to-video
Kling v3 Omni at 720P. Multi-image reference video generation — supply up to 4 images and reference them in your prompt with <<<image_N>>>. Apimart-backed.
kling-v3.0-omni-pro-image-to-video
Kling v3 Omni at 1080P. Multi-image reference video generation — supply up to 4 images and reference them in your prompt with <<<image_N>>>. Apimart-backed.
kling-v3.0-omni-4k-image-to-video
Kling v3 Omni at 4K. Multi-image reference video generation — supply up to 4 images and reference them in your prompt with <<<image_N>>>. Apimart-backed.
kling-v3-turbo-pro-image-to-video
Generate fast, high-quality videos from a single image using Kling v3 Turbo Pro (1080p). Supports durations from 3 to 15 seconds.
vidu-q3-pro-image-to-video
Vidu Q3 Pro Image-to-Video animates a single starting image into a smooth, prompt-guided clip up to 1080p. It preserves character identity, lighting, and composition while introducing natural motion, camera moves, and atmosphere — ideal for bringing concept art, product shots, and stills to life.
kling-v3-turbo-standard-image-to-video
Generate fast, high-quality videos from a single image using Kling v3 Turbo Standard (720p). Supports durations from 3 to 15 seconds.
seedance-2.1-image-to-video
Seedance 2.1 Image-to-Video converts a single still image into a high-fidelity cinematic video. This next-generation model delivers enhanced temporal consistency, improved motion realism, and smarter prompt adherence — producing smooth, coherent clips up to 1080p.
seedance-2.5-image-to-video
Seedance 2.5 Image-to-Video is the early-access preview build of the Seedance 2.5 family, animating a single image into video. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview).
flux-3-image-to-video
FLUX 3 Image-to-Video animates a still image into a cinematic clip with optional native synchronized audio, using Black Forest Labs' unified image/video/audio architecture. Motion stays physically grounded and consistent with the source frame, making it suited for product animation, portrait bring-to-life effects, and scene extension.
happy-horse-1.1-image-to-video-1080p
Happy Horse 1.1 Image to Video (1080p) — bring still images to life with fluid, expressive 1080p animation.
happy-horse-1.1-image-to-video-720p
Happy Horse 1.1 Image to Video (720p) — bring still images to life with fluid, expressive 720p animation.
happy-horse-1.1-reference-to-video-720p
Happy Horse 1.1 Reference to Video (720p) — generate 720p video conditioned on 1-9 reference images plus a text prompt.
happy-horse-1.1-reference-to-video-1080p
Happy Horse 1.1 Reference to Video (1080p) — generate 1080p video conditioned on 1-9 reference images plus a text prompt.
runway-image-to-video
Animate any image by turning it into a video with motion effects or scene continuity. RunwayML’s I2V model transforms static visuals into short clips by extrapolating depth, movement, and temporal dynamics.
seedance-2-vip-image-to-video-4k
SD 2 Image-to-Video VIP 4K by ByteDance. Animates a still image into a 4K ultra-HD video with priority routing and 4–15 second duration.
seedance-2-vip-first-last-frame-4k
SD 2 First & Last Frame VIP 4K by ByteDance. Generate 4K video that transitions between two reference images with priority routing. Provide 1 image for start-frame-only, or 2 images for both start and end frames.
seedance-2-vip-omni-reference-4k
SD 2 Omni Reference VIP 4K by ByteDance. Generate 4K ultra-HD videos using up to 9 image references, up to 3 video clips, and up to 3 audio references with priority routing. Reference materials in your prompt with @image1…@image9, @video1…@video3, and @audio1…@audio3.
veo3-fast-image-to-video
Quickly transform static images into short, motion-rich video clips with fast rendering and impressive quality — powered by Google's VEO3 on MuAPI.
vidu-q1-reference
Vidu Q1 enables you to generate cinematic 1080p videos using multiple visual references—up to seven images—and text prompts. Designed for consistency, it preserves character appearance, props, and backgrounds across scenes while adding new motion and narrative elements.
minimax-hailuo-02-pro-i2v
Advanced image-to-video with cinematic realism. Adds dynamic camera motion, realistic physics, and atmospheric detail for storytelling.
video-effects
AI Video Effects applies advanced visual transformations, color grading, and cinematic filters to create stunning videos from images.
seedance-lite-i2v
Seedance Lite I2V version animates static images into short videos quickly, focusing on basic motion effects and efficient processing—best suited for fast demos or mobile-friendly use.
pixverse-v5-i2v
PixVerse V5 delivers a major leap forward in AI-powered video creation — now featuring smoother motion, ultra-high resolution, and expanded visual effects.
seedance-2.5-spicy-image-to-video
Seedance 2.5 Spicy Image-to-Video is the relaxed-moderation variant of the Seedance 2.5 flagship model. It animates a single image into photorealistic 4K video with reduced content-safety filtering and more dramatic, higher-contrast motion than the standard tier — while keeping native audio generation and precise camera trajectory control.
pixverse-v4.5-i2v
Upload an image and PixVerse v4.5 will breathe life into it with smooth camera motion, realistic effects, and animated elements. Whether it’s a portrait, landscape, or concept art, this mode turns still visuals into dynamic short videos.
wan2.1-image-to-video
Animate static images into expressive video sequences with WAN 2.1. Upload any image and guide its transformation into a moving scene — great for bringing art, characters, or photos to life with smooth motion and consistent style.
seedance-2-spicy-image-to-video
Seedance 2 Spicy Image-to-Video by ByteDance. Animates a start frame into cinematic video with VIP-tier priority routing and up to 2K resolution, with reduced content-safety filtering for more creative freedom.
seedance-2-spicy-image-to-video-fast
Seedance 2 Spicy Image-to-Video Fast by ByteDance. The quickest Spicy-tier image animation, with reduced content-safety filtering and the same fast queue as Seedance 2 VIP Fast.
wan2.5-image-to-video
WAN 2.5 Image-to-Video takes your image as the starting frame and turns it into a dynamic video, preserving realism, motion, and camera effects. Upload a static image, add a descriptive text prompt, and the model generates cinematic motion—camera pans, environmental movement, and realistic physics—across the result.
wan2.5-image-to-video-fast
Convert a single static image into a cinematic short video with realistic motion, dynamic camera movement, and environmental effects. The Fast mode generates high-quality videos quickly, perfect for rapid prototyping, social media clips, and immersive visual storytelling from still images.
minimax-h3-reference-to-video
MiniMax H3 Reference to Video creates a 2K video from a prompt plus image, video, and optional audio references through the Muapi API.
wan3.0-image-to-video
Wan 3.0 Image to Video animates a source image with synchronized audio from a motion prompt, with an optional end-frame image, selectable resolution, aspect ratio, and duration.
veo3.1-image-to-video
Veo 3.1 is Google's advanced AI video generation model that allows users to create high-quality, 8-second videos from static images. This feature is particularly useful for transforming concept art, storyboards, or static visuals into dynamic video clips with synchronized audio.
seedance-lite-reference-video
Seedance Lite's Reference-to-Video feature allows you to supply up to 4 images as reference inputs. The model intelligently blends aspects from these images to generate a cohesive, high-quality video.
minimax-h3-image-to-video
MiniMax H3 Image to Video animates a source image with a motion prompt through the Muapi API.
ltx-2-fast-image-to-video
LTX-2 Fast is a speed-optimized mode of the LTX-2 engine by Lightricks, focused on generating short video clips from a still image + prompt (I2V) with good fidelity and rapid turnaround. It supports audio/video together, multiple aspect ratios, and is ideal when you need quick output for iteration or storyboarding.
vidu-q2-turbo-start-end-video
Vidu Q2 Turbo Start–End Video creates highly detailed cinematic sequences by interpolating between two visual states — your start frame and end frame. Built for story moments, cinematic transformations, product reveals, and artistic transitions, it captures smooth motion, realistic lighting shifts, and dynamic camera movements while maintaining fidelity and emotional tone.
vidu-q2-pro-start-end-video
Vidu Q2 Pro Start–End Video is a professional-grade model built for cinematic transformation storytelling. It evolves a scene, subject, or concept from one moment to another through smooth visual interpolation, natural lighting transitions, and dynamic motion.
seedance-2.5-image-to-video-480p
Seedance 2.5 Image-to-Video 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, animating a single image into video. Faster and more cost-effective than the 720p variant. Supports clips up to 30 seconds.
seedance-2.5-omni-reference-480p
Seedance 2.5 Omni Reference 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, blending multiple reference images, video clips, and audio into a single guided generation. Faster and more cost-effective than the 720p variant.
kling-o1-image-to-video
Kling O1’s Image-to-Video mode transforms one or more reference images into short cinematic video clips by adding natural motion, camera choreography, and scene dynamics while preserving subject identity and visual consistency. It supports start/end frames.
ltx-2-pro-image-to-video
LTX-2 Pro is the high-fidelity video-generation engine by Lightricks designed for professional workflows, supporting both text-to-video and image-to-video inputs. It enables realistic motion, synchronized audio-video, cinematic camera moves and stylized visuals. Ideal for your timeline-based video interface: you supply a prompt or image, define duration/aspect ratio, then it generates a clip that you can ingest, rename, batch-move, split or timeline-edit.
veo3.1-reference-to-video
Veo 3.1 R2V allows creators to generate dynamic videos using up to three reference images. The model maintains visual consistency of characters, objects, and style throughout the video, producing cinematic-quality 8-second clips. It’s perfect for turning concept art, storyboards, or character designs into short, animated sequences while preserving original aesthetics.
seedance-2.5-first-last-frame-480p
Seedance 2.5 First & Last Frame 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, generating a smooth video transition between a start and end image. Faster and more cost-effective than the 720p variant.
kling-o1-standard-reference-to-video
Kling O1 Standard Reference-to-Video generates a smooth, realistic video using one or multiple reference images as visual guidance. It preserves the visual identity, composition, and lighting from the references while adding subtle camera motion, natural parallax, and light environmental animation. This mode prioritizes stability and realism, making it ideal for character shots, environments, product visuals, and calm cinematic scenes.
seedance-2.5-image-to-video-1080p
Seedance 2.5 Image-to-Video is the early-access preview build of the Seedance 2.5 family, animating a single image into video. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier.
minimax-h3-open-image-to-video
MiniMax H3 Open Image to Video animates a first-frame image (with optional last-frame guidance) into coherent video with native audio, supporting 480p/768p resolution and 5-15s duration.
minimax-h3-open-reference-to-video
MiniMax H3 Open Reference to Video generates coherent videos from prompts and multimodal references, guided by up to 9 reference images, 3 reference videos, and 3 reference audios with native stereo audio.
ltx-2.3-image-to-video
LTX-2.3 Image-to-Video animates a single image into a coherent cinematic clip. It preserves scene composition and lighting while adding smooth camera motion, parallax, and environmental dynamics. Built on the upgraded LTX-2.3 architecture for sharper output and improved temporal consistency.
seedance-2.5-omni-reference-4k
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier.
seedance-2.5-spicy-first-last-frame
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview). This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
wan3.0-spicy-image-to-video
Wan 3.0 Spicy Image to Video animates a source image into a bold, high-contrast, high-motion video with synchronized audio from a motion prompt, with an optional end-frame image, selectable resolution, aspect ratio, and duration.
seedance-2.5-spicy-first-last-frame-1080p
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
seedance-2.5-image-to-video-4k
Seedance 2.5 Image-to-Video is the early-access preview build of the Seedance 2.5 family, animating a single image into video. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier.
seedance-2.5-first-last-frame-1080p
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier.
seedance-2.5-first-last-frame-4k
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier.
seedance-2.5-omni-reference-1080p
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier.
seedance-2-omni-reference-480p
SD 2.0 480p Omni Reference — generate videos with visual consistency using reference images, videos, and audio at 480p resolution. More cost-effective than the 720p variant. Supports up to 9 images, 3 video clips, and 3 audio clips. Use @image1, @video1, @audio1 syntax in your prompt.
seedance-2.5-intl-image-to-video
Seedance 2.5 Image-to-Video is the early-access preview build of the Seedance 2.5 family, animating a single image into video. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview). This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
pixverse-v6-transition
Create a smooth transition between two images (start and end) or from a single starting image to a generated video.
seedance-2.5-intl-image-to-video-480p
Seedance 2.5 Image-to-Video 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, animating a single image into video. Faster and more cost-effective than the 720p variant. Supports clips up to 30 seconds. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
seedance-2.5-spicy-first-last-frame-4k
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
seedance-2.5-intl-omni-reference-480p
Seedance 2.5 Omni Reference 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, blending multiple reference images, video clips, and audio into a single guided generation. Faster and more cost-effective than the 720p variant. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
seedance-2-vip-image-to-video
SD 2 Image-to-Video VIP (Pro) by ByteDance. Animates a start-frame image into a high-quality video with priority routing, native audio, 4–15 second duration, and 2K resolution.
seedance-2.5-spicy-image-to-video-4k
Seedance 2.5 Image-to-Video is the early-access preview build of the Seedance 2.5 family, animating a single image into video. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
seedance-2.5-intl-first-last-frame-480p
Seedance 2.5 First & Last Frame 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, generating a smooth video transition between a start and end image. Faster and more cost-effective than the 720p variant. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
seedance-2-vip-first-last-frame
SD 2 First & Last Frame VIP (Pro) by ByteDance. Generate video that transitions between two reference images with priority routing. Provide 1 image for start-frame-only, or 2 images for both start and end frames.
seedance-2-vip-first-last-frame-fast
SD 2 First & Last Frame VIP Fast by ByteDance. Faster generation of video transitions between two reference images with priority routing.
seedance-2.5-spicy-image-to-video-480p
Seedance 2.5 Image-to-Video 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, animating a single image into video. Faster and more cost-effective than the 720p variant. Supports clips up to 30 seconds. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
seedance-2-vip-omni-reference
SD 2 Omni Reference VIP (Pro) by ByteDance. Generate videos using up to 9 image references, up to 3 video clips, and up to 3 audio references with priority routing. Reference materials in your prompt with @image1…@image9, @video1…@video3, and @audio1…@audio3. Also supports @omni-character:<char_id> for trained characters.
seedance-2.5-spicy-image-to-video-1080p
Seedance 2.5 Image-to-Video is the early-access preview build of the Seedance 2.5 family, animating a single image into video. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
seedance-2-vip-omni-reference-1080p
SD 2 Omni Reference VIP 1080p by ByteDance. Generate full HD videos using up to 9 image references, up to 3 video clips, and up to 3 audio references with priority routing. Reference materials in your prompt with @image1…@image9, @video1…@video3, and @audio1…@audio3.
vidu-q3-pro-first-last-frames
Vidu Q3 Pro First-Last Frames interpolates a smooth, cinematic transition between two key images — your start frame and end frame — guided by a text prompt. Perfect for transformation reveals, scene transitions, product morphs, and storytelling beats that need a clean, controlled arc from A to B.
seedance-2.5-spicy-omni-reference
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview). This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
vidu-q3-turbo-image-to-video
Vidu Q3 Turbo Image-to-Video animates a starting image into a fast, prompt-guided clip while keeping subject identity and composition intact. Built for speed and cost efficiency — perfect for batch animation, social content, and quick creative exploration.
vidu-q2-pro-image-to-video
Vidu Q2 Pro Image-to-Video animates a single starting image into a smooth, prompt-guided clip up to 1080p while preserving subject identity, lighting, and composition.
seedance-2.5-intl-first-last-frame
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview). This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
vidu-q3-turbo-first-last-frames
Vidu Q3 Turbo First-Last Frames interpolates a quick, cost-efficient transition between two key images — your start frame and end frame — guided by a text prompt. Great for transformation reveals, transitions, and short-form storytelling at scale.
seedance-2.5-spicy-first-last-frame-480p
Seedance 2.5 First & Last Frame 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, generating a smooth video transition between a start and end image. Faster and more cost-effective than the 720p variant. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
seedance-2.5-intl-first-last-frame-1080p
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
wan2.6-image-to-video-spicy
Alibaba WAN 2.6 Spicy Image-to-Video animates a still image into a bold, high-motion clip with optional audio-guided generation, single- or multi-shot framing, and prompt-expansion.
seedance-2.5-intl-first-last-frame-4k
Seedance 2.5 First & Last Frame is the early-access preview build of the Seedance 2.5 family, generating a smooth video transition between a start and end image. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
flux-3-omni
FLUX 3 Omni is Black Forest Labs' Omni Reference mode — it takes multiple reference images and video clips at once and generates new images or video that keep the same subject, style, and scene consistent across every output. Listed by Black Forest Labs as coming soon.
seedance-2.5-intl-omni-reference
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. Supports clips up to 30 seconds and 480p/720p output (1080p is not yet supported by this preview). This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
seedance-2.5-spicy-omni-reference-480p
Seedance 2.5 Omni Reference 480p is the early-access preview build of the Seedance 2.5 family at 480p resolution, blending multiple reference images, video clips, and audio into a single guided generation. Faster and more cost-effective than the 720p variant. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
seedance-2.5-intl-omni-reference-1080p
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
seedance-2.5-spicy-omni-reference-1080p
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. This 1080p variant is natively rendered by Seegen (not upscaled from 720p) and supports clips up to 30 seconds. It is priced above the standard 720p tier. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
seedance-2.5-intl-omni-reference-4k
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier. This international-region endpoint is served via Seegen's Dreamina-hosted deployment of the same Seedance 2.5 model, for traffic outside mainland China.
seedance-2.5-spicy-omni-reference-4k
Seedance 2.5 Omni Reference is the early-access preview build of the Seedance 2.5 family, blending multiple reference images, video clips, and audio into a single guided generation. This 4K variant is upscaled from the model's native 720p render (not a native 4K output) and supports clips up to 30 seconds. Upscaling to 4K costs more than the standard 720p tier. This Spicy endpoint is the relaxed-moderation sibling of the standard tier, with lighter content-safety filtering and bolder, higher-contrast output.
wan2.7-image-to-video-spicy
Alibaba WAN 2.7 Spicy Image-to-Video animates a still image into a bold, high-motion clip with optional audio-guided generation, single- or multi-shot framing, and prompt-expansion.
wan3.0-reference-to-video
Wan 3.0 Reference to Video generates a video guided by a prompt plus up to 10 reference images, 5 reference videos, and 5 reference audios for visual and audio coherence.
wan3.0-prime-image-to-video
Wan 3.0 Prime Image to Video animates a source image with synchronized audio from a motion prompt at higher fidelity, with an optional end-frame image, selectable resolution, aspect ratio, and duration.
ltx-2.5-image-to-video
LTX 2.5 Image-to-Video animates a first-frame image into high-fidelity synchronized audio-video content, with optional last-frame guidance and 720P / 1080P / 2K / 4K output for cinematic videos, social content, ads, and production workflows.
wan3.0-spicy-reference-to-video
Wan 3.0 Spicy Reference to Video generates a bold, high-contrast, high-motion video guided by a prompt plus up to 10 reference images, 5 reference videos, and 5 reference audios for visual and audio coherence.
ai-video-effects
AI Video Effects applies advanced visual transformations, color grading, and cinematic filters to create stunning videos from images.
motion-controls
Motion Controls adds dynamic camera movements, speed ramps, and zoom effects to bring your images to life as smooth, engaging videos.
vfx
VFX delivers high-impact visual effects like explosions, particles, and cinematic overlays to transform static images into action-packed videos.
常见问题
What image formats are supported?
Pass a public URL to a JPEG, PNG, or WebP image — or upload via `POST /api/v1/upload_file` first to get a hosted URL.
How do I keep the subject consistent across multiple shots?
Use the same reference image across multiple image-to-video calls with different prompts. For tighter consistency, use the workflow builder to chain a face-preservation node between clips.
Which model handles complex camera moves best?
Kling Pro and Veo 3 handle dolly, pan, and orbit moves most reliably. Pass camera instructions in the prompt — most models parse English camera terminology.