Cookie Preferences
We use cookies to personalize content and analyze traffic. See our Cookie Policy.
还没有视频。提交表单以生成视频。
透明定价，无隐藏费用。按需支付。
| 规则与模式 | 渠道 | 积分 | 价格（美元） | 官方/参考价格 | 每日节省 |
|---|---|---|---|---|---|
720p-video videoKlingmode: 720p | Regular | 24per second | $0.109 | Fal$0.126per second | 按每日 1,000 个 10 秒视频估算用 Fal 原价每天多花 $168.950ApiPass 低 13.41% |
1080p-video videoKlingmode: 1080p | Regular | 32per second | $0.145 | Fal$0.168per second | 按每日 1,000 个 10 秒视频估算用 Fal 原价每天多花 $225.450ApiPass 低 13.42% |
Kling-3-motion-control 使用完整指南
在 ApiPass 上访问 Kling 3.0 Motion Control API，将参考视频中的动作迁移到角色图像，同时保留面部身份、表情和真实的动作动态。结合图像与视频参考、可选的提示词引导，以及 720p 或 1080p 输出模式，构建一致的 AI 视频动画。
Kling Motion Control 3.0 是 Kuaishou Kling 面向视频到视频动作迁移的 AI 视频动作模型。它接收一张展示目标角色的参考图像和一段定义动作模式的参考视频，然后生成一个新视频，让角色跟随参考动作，同时保持稳定的面部身份、自然的表情和真实的运动效果。与 Kling 2.6 Motion Control 相比，Kling VIDEO 3.0 Motion Control 在复杂场景、多角度镜头、动态构图和更长动作序列中的面部一致性与动作稳定性方面都有提升。
通过 input_urls 提供角色图像，并通过 video_urls 提供动作参考视频。Kling 3.0 Motion Control 会将视频中的动作、手势和表演迁移到角色图像上，同时保持生成视频的连贯性。
该模型旨在在转身、镜头变化和动态运动过程中保留面部身份。这使其适合用于为肖像、插画角色、头像、吉祥物和虚拟主持人制作外观一致的动画。
使用可选的 prompt 字段来引导期望的输出、动画内容、风格或场景行为。提示词可以为空，也可以长达 2500 个字符，让开发者能够以轻量方式控制动作驱动生成的创意方向。
Kling 3.0 Motion Control 可在不同镜头角度和长动作序列中保持面部身份稳定。角色在转身、移动或执行复杂动作时仍能维持相同的面部结构，使该 API 非常适合生成一致的 AI 视频。
Kling Motion Control 3.0 能准确捕捉微妙的情绪变化，例如微笑、惊讶或悲伤。开发者可以生成富有表现力的 AI 视频，在从参考视频迁移动作的同时保持情绪自然。
即使在运动或镜头切换过程中面部部分区域短暂被遮挡，Kling VIDEO 3.0 Motion Control 也能保留面部结构。生成的视频会平滑恢复面部细节，同时保持身份一致性。
当镜头构图变化或出现动态运动时，Kling 3.0 Motion Control 仍能保持面部清晰。通过动作控制流程创建的缩放、平移和电影感镜头中，面部依然清晰且易于识别。
使用真人动作驱动插画角色、头像或吉祥物。通过结合参考视频和角色图像，Kling 3.0 Motion Control 可迁移自然的身体动作和手势，创建逼真的动画表演。
将热门舞蹈、手势或表情应用到 AI 角色上，生成短视频。创作者、网红和品牌可以快速制作适用于 TikTok、YouTube Shorts 以及其他社交平台的吸引人内容。
将静态视觉素材、吉祥物或虚拟角色转化为动态宣传视频。品牌可以使用参考素材中的动作来创建主持人、产品演示和品牌化表演片段。
在正式制作前快速原型化场景和角色表演。影视创作者和创意团队可以借助动作控制的 AI 视频生成，测试编舞、表演、镜头运动和角色一致性。
使用清晰且准备充分的图像和视频参考，以提升生成视频中的身份一致性、动作准确性和情绪表现质量。
使用清晰、光线充足且能明确看到面部的角色图像。该 API 依赖面部数据来保持身份一致性，因此高质量的面部参考有助于提升稳定性和动作准确性。
上传与最终希望生成效果高度接近的面部参考。动作控制流程非常重视面部结构，因此准确的参考有助于更可靠地保留身份。
为了获得更真实的头部运动，请使用覆盖正面、侧面等有用视角的参考素材。更多面部信息可以改善角度过渡和转头准确性。
在生成复杂表演或情绪变化时，选择质量较高的动作参考视频。视频能够提供更丰富的动作和面部信息，帮助 Kling VIDEO 3.0 Motion Control 创建更流畅、更一致的结果。
ApiPass 为 Kling 3.0 Motion Control 提供了直观的基于任务的工作流。使用 POST /api/v1/jobs/createTask 创建任务，然后通过 GET /api/v1/jobs/recordInfo?taskId=<taskId> 获取结果，或使用可选的 callBackUrl 接收完成通知。
使用必需的图像和视频输入、可选的提示词引导、720p 或 1080p 输出模式，以及通过 character_orientation 支持的参考朝向设置，构建稳定的动作控制视频功能。
使用最大 10MB 的 JPEG、PNG 或 JPG 图像，并确保主体清晰可见；动作视频需为 3 到 30 秒的 MP4 或 QuickTime 格式。输入图像应清楚展示主体，并满足支持的 2:5 到 5:2 宽高比范围。
从 ApiPass Playground 开始，快速测试 Kling 3.0 Motion Control。上传参考角色图像和动作参考视频，添加可选提示词，选择角色朝向和输出模式，然后在集成 API 前预览动作迁移结果。
创建你的 ApiPass API key，并为 input_urls 和 video_urls 准备可公开访问的文件 URL。图像应为 JPEG、PNG 或 JPG，视频应为 MP4 或 QuickTime。检查 prompt、mode、character_orientation 以及可选 callBackUrl 等请求参数。
使用 model kling/kling-v3-motion-control 向 https://api.apipass.dev/api/v1/jobs/createTask 发送 POST 请求。响应会返回 taskId，你可以通过 https://api.apipass.dev/api/v1/jobs/recordInfo?taskId=YOUR_TASK_ID 进行查询，直到生成的视频结果可用。
所有 API 都需要通过 Bearer Token 进行身份验证。
Authorization: Bearer
Create an asynchronous video-to-video task for kling/kling-v3-motion-control. This model transfers motion from a reference video to a character image while preserving facial identity, expressions, and realistic motion dynamics.
1{
2 "model": "kling/kling-v3-motion-control",
3 "channel": "auto",
4 "callBackUrl": "https://your-domain.com/api/callback",
5 "input": {
6 "character_orientation": "video",
7 "input_urls": [
8 "https://static.aiquickdraw.com/tools/example/1767694885407_pObJoMcy.png"
9 ],
10 "mode": "720p",
11 "prompt": "The cartoon character is dancing.",
12 "video_urls": [
13 "https://static.aiquickdraw.com/tools/example/1767525918769_QyvTNib2.mp4"
14 ]
15 }
16}model必填stringApiPass model endpoint name. Use kling/kling-v3-motion-control for this endpoint.
kling/kling-v3-motion-control
channel可选stringDefault channel is auto. APIPASS automatically allocates tasks across available providers based on real-time pricing and stability metrics. starter is ultra-low-cost with limited quotas and weaker stability; regular is standard and much cheaper than official APIs with moderate stability; official uses the model native API with high stability, fast task execution, and official pricing.
Default channel is auto. APIPASS automatically allocates tasks across available providers based on real-time pricing and stability metrics. starter is ultra-low-cost with limited quotas and weaker stability; regular is standard and much cheaper than official APIs with moderate stability; official uses the model native API with high stability, fast task execution, and official pricing.
可用选项：
auto
callBackUrl可选stringOptional callback URL. If provided, ApiPass sends task status and result updates to this URL when the asynchronous task completes or fails.
https://your-domain.com/api/callback
prompt可选stringText prompt used to guide the generated animation content. Can be empty. Maximum length is 2500 characters.
The cartoon character is dancing.
input_urls必填array<string>Reference image URL. The generated video uses the character, background, and other visual elements from this image. Upload the image first and provide its URL here.
Use a clear image showing the subject's head, shoulders, and torso. Supported formats are JPEG, PNG, or JPG. Maximum file size is 10MB, image size should be greater than 340px, and aspect ratio should be between 2:5 and 5:2.
[ "https://static.aiquickdraw.com/tools/example/1767694885407_pObJoMcy.png" ]
video_urls必填array<string>Reference video URL. The generated video's character actions follow the motion pattern from this video. Upload the video first and provide its URL here.
Supported formats are MP4 and QuickTime. Reference video duration must be between 3 and 30 seconds; when character_orientation is image, use a video up to 10 seconds, and when character_orientation is video, use a video up to 30 seconds.
[ "https://static.aiquickdraw.com/tools/example/1767525918769_QyvTNib2.mp4" ]
mode可选stringOutput resolution mode. Use 720p for standard mode or 1080p for professional mode.
可用选项：
720p
character_orientation可选stringReference source for character orientation. video keeps the generated character orientation consistent with the reference video and is recommended. image keeps the generated character orientation consistent with the person in the reference image. Default value is video.
可用选项：
video
1curl -X POST "https://api.apipass.dev/api/v1/jobs/createTask" \
2 -H "Authorization: Bearer YOUR_API_KEY" \
3 -H "Content-Type: application/json" \
4 -d '{
5 "model": "kling/kling-v3-motion-control",
6 "channel": "auto",
7 "input": {
8 "character_orientation": "video",
9 "input_urls": [
10 "https://static.aiquickdraw.com/tools/example/1767694885407_pObJoMcy.png"
11 ],
12 "mode": "720p",
13 "prompt": "The cartoon character is dancing.",
14 "video_urls": [
15 "https://static.aiquickdraw.com/tools/example/1767525918769_QyvTNib2.mp4"
16 ]
17 }
18}'1{
2 "taskId": "task_xxx",
3 "status": "queued"
4}taskIdUnique task ID. Use this value to query task status and results with the recordInfo endpoint.
statusInitial task status after creation, such as queued.
ByteDance
ByteDance
起价
0 积分
ByteDance
ByteDance
起价
0 积分
ByteDance
ByteDance
Turn text, images, video clips, and audio into cinematic, audio-synced videos with a single API call.
起价
0 积分
MiniMax
MiniMax
起价
0 积分
ByteDance
ByteDance
Generate cinematic 4K AI videos up to 30 seconds with Seedance 2.5's multimodal power.
起价
0 积分
Google Veo 3.1 is the latest state-of-the-art generative video model designed to transform text and image prompts into high-fidelity cinematic visuals. Building upon its predecessors, Veo 3.1 features significant upgrades in prompt adherence, visual realism, and native audio generation, creating 8-second clips with synchronized sound effects and dialogue.
起价
0 积分
wan
wan
起价
0 积分
Runway
Runway
起价
0 积分
Kling
Kling
Transform text prompts and images into high-quality videos with advanced AI. Generate cinematic content with customizable duration, aspect ratio, and audio capabilities.
起价
0 积分
ByteDance
ByteDance
Generate realistic human animation videos by combining a portrait image with audio, producing natural lip-sync, facial expressions, and body movements.
起价
0 积分
MiniMax
MiniMax
Generate high-quality AI videos from text prompts or images with MiniMax's Hailuo 2.3 model. Support for both 6s and 10s durations, 768p and 1080p resolutions, with intelligent prompt optimization.
起价
0 积分
grok
grok
Grok Imagine is xAI’s AI video generation model for creating high-quality videos with realistic motion, native audio, and synchronized speech. Grok Imagine 1.5 introduces faster generation speeds, improved motion physics, and stronger visual consistency for modern AI video applications.
起价
0 积分
Luma
Luma
起价
0 积分
Kling
Kling
Kling v3 video generation model, supports text-to-video and image-to-video, up to 15 seconds, 1080p pro mode
起价
0 积分
Kling
Kling
Generate realistic talking avatar videos from a single image and audio file using Kling AI Avatar v2 API. Create lifelike lip-sync animations with natural expressions and body movements.
起价
0 积分
起价
0 积分
kling
kling
起价
0 积分