Actions24
- Audio Actions
- Image Actions
- Video Actions
- Animate Mix | 视频换人
- Animate Move | 图生动作
- Emoji Video | Emoji 视频生成
- Image to Video (First & Last Frames) | 图生视频(首尾帧)
- Image to Video (First Frame) | 图生视频(首帧)
- Video Continuation | 视频续写
- Reference to Video | 参考生视频(wan2.6)
- Text to Video | 文生视频
- Video Editing & Effects | 视频编辑与特效
- Video Retalk (Lip Sync) | VideoRetalk 口型复刻
- Task Actions
- Vision Actions
- File Actions
Video → Image to Video (First & Last Frames) | 图生视频(首尾帧)
AI-generatedSummary
Generate a video clip that smoothly transitions between a specified first frame and last frame using an AI model optimized for start-end frame constraints.
Inputs
- Model (required) — Select the AI model variant for generating the video, tuned for start-end frame constraints (options: Wan 2.7 Image-to-Video, Wan2.2 Keyframe-to-Video Flash, Wanx2.1 Keyframe-to-Video Plus).
- First Frame URL/Base64 (required) — Provide the starting keyframe as an image URL or base64 data URI.
- Last Frame URL/Base64 (required) — Provide the ending keyframe as an image URL or base64 data URI.
- Prompt — Add a narrative prompt to guide the artistic blending of the two keyframes.
- Negative Prompt — Specify undesired elements to suppress in the generated video clip.
- Effect Template — Optionally specify a motion template to align with the keyframe content.
- Audio URL — For model 'wan2.7-i2v' only: provide an optional driving audio URL to influence generation.
- Resolution — Set the resolution tier for keyframe interpolation; available options depend on the chosen model (e.g., 480P or 720P for some models, 720P or 1080P for wan2.7-i2v).
- Duration (Seconds) — For model 'wan2.7-i2v' only: set an integer duration from 2 to 15 seconds for the video length.
- Auto Enhance Prompt — Enable or disable automatic rewriting of prompts for smoother transitions by Aliyun Bailian.
- Advanced Options — Configure additional settings such as embedding a watermark, specifying a deterministic seed, or enabling runtime seed randomization.
- Wait for Task Completion — Choose whether to wait for the AI generation task to complete by polling or return immediately with the task ID.
- Poll Interval (MS) — Specify delay between polling attempts if waiting for task completion, in milliseconds.
- Max Wait (Seconds) — Set an upper time limit for waiting for task completion before timing out.
Output shape
a single video generation task response or, if wait for task is enabled, the completed task result including the generated video URL
If 'Wait for Task Completion' is true, the node polls the task status and outputs the final video data upon success or an error upon failure. If false, it returns immediately with the task ID for manual tracking. Optional parameters and resolution vary depending on chosen AI model variant.
Examples
Example 1: Generate a 5-second 1080p video transitioning between two specified keyframes using the wan2.7-i2v model with automatic prompt enhancement and wait for completion enabled.
Model='wan2.7-i2v', First Frame URL='...', Last Frame URL='...', Prompt='...', Duration=5, Resolution=1080P, Auto Enhance Prompt=true, Wait for Task Completion=true