Wan2.7: Text to Video
Generate videos from text prompts using Wan 2.7. Supports audio reference to create lip-synced video
About
This ComfyUI workflow turns a plain text prompt into a short video using the Wan 2.7 model. At its core, the Wan2TextToVideoApi node sends your prompt and generation settings to the Wan 2.7 backend, which synthesizes a sequence of frames that match the described scene, style, and motion. If you supply a reference audio clip, the node uses it to guide mouth shapes and timing, enabling convincing lip-synced results without manual keyframing.
What you provide
- Reading the workflow
Runs in ComfyUI. This workflow can't run on this page. Open it in Comfy Cloud with everything loaded, or download it for your own ComfyUI.
Try in CloudQuestions & answers
If your node targets a hosted Wan 2.7 endpoint, you will typically need to provide an API key or credentials in the node’s settings. Check the node’s documentation or your service provider for details.
Clip length depends on the limits configured in the Wan2TextToVideoApi node and the backend service. This workflow is optimized for short clips; longer durations increase generation time and compute cost.
Use clean, speech-focused audio (avoid music/noise), trim leading silence, keep the voice centered and at a consistent level, and match the language of your text prompt to the spoken audio. Ensure the audio sample rate and format are supported by the node.
Yes. Set a fixed seed in Wan2TextToVideoApi (if exposed). With a fixed seed and identical settings, results will be more consistent between runs; without it, outputs may vary.