Wan 2.7 video edit

Edit a source video using text instructions and two reference images, and return one edited video. Upload the video and both reference images. Reference image 1 connects to image1; reference image 2 connects to image2. Use those names in your instructions to explain how each reference should affect the edit. Both image uploads are required by this saved Pipe. For a first run with the default instructions, upload a clip of two people playing chess, a bear as reference image 1 and a dog as reference image 2. For other footage, rewrite the instructions to describe the changes and details to retain. Requests to preserve the scene or movement guide the model; they do not guarantee unchanged pixels or motion. Start with the defaults: 720P, 16:9 and duration auto. Choose 720P or 1080P and the aspect ratio for your intended output. Duration auto matches the input video duration; choosing 2–10 seconds truncates from the beginning of the source video. Audio mode auto lets the model decide whether to regenerate audio based on your instructions. Include desired sounds in the instructions when using auto. Choose origin to preserve the input audio; requests for new sounds should then not be expected to take effect. Seed -1 randomizes each run. A nonnegative seed controls generation variation. Enable Watermark to request the provider's AI-generated watermark. Under Video encoding, leave file type and codec on auto for automatic selection. Supported file types are mp4, mkv and webm; supported codecs are h264 and av1. This Pipe uses the Wan 2.7 partner API and requires configured provider access. Uploaded video, reference images and instructions are sent to the provider, and generation uses a provider-priced API call.

Edit a source video using text instructions and two reference images, and return one edited video. Upload the video and both reference images. Reference image 1 connects to image1; reference image 2 connects to image2. Use those names in your instructions to explain how each reference should affect the edit. Both image uploads are required by this saved Pipe. For a first run with the default instructions, upload a clip of two people playing chess, a bear as reference image 1 and a dog as reference image 2. For other footage, rewrite the instructions to describe the changes and details to retain. Requests to preserve the scene or movement guide the model; they do not guarantee unchanged pixels or motion. Start with the defaults: 720P, 16:9 and duration auto. Choose 720P or 1080P and the aspect ratio for your intended output. Duration auto matches the input video duration; choosing 2–10 seconds truncates from the beginning of the source video. Audio mode auto lets the model decide whether to regenerate audio based on your instructions. Include desired sounds in the instructions when using auto. Choose origin to preserve the input audio; requests for new sounds should then not be expected to take effect. Seed -1 randomizes each run. A nonnegative seed controls generation variation. Enable Watermark to request the provider's AI-generated watermark. Under Video encoding, leave file type and codec on auto for automatic selection. Supported file types are mp4, mkv and webm; supported codecs are h264 and av1. This Pipe uses the Wan 2.7 partner API and requires configured provider access. Uploaded video, reference images and instructions are sent to the provider, and generation uses a provider-priced API call.