Wan 2.7 (BudgetPixel)

Wan 2.7 (video) — Alibaba's video model, with a driving-audio track

By budgetpixel-ai·Created about 17 hours ago·Updated about 12 hours ago· 3

audio
end_image
image
reference_images
video

video

◄prompt►

◄aspect_ratio16:9►

◄generate_audiotrue►

◄length_seconds5►

◄resolution720p►

◄force_rerunfalse►

Watch the category before you add this one - the pack has two nodes displayed as "Wan 2.7", and this is the video model under BudgetPixel/Video, not the image one. It's Alibaba's Wan 2.7 video tier: text-to-video, image-to-video with a start and optional end frame, and video-to-video editing, all at 720p or 1080p in 5/10/15-second clips. At 100 credits per second it sits in the affordable middle of the pack's video line - cheaper than Seedance 2.5's 330, pricier than SeeDance Mini's 120 - and it defaults to audio on (generate_audio defaults true), which most of the other video nodes don't.

What's interesting about it

The driving-audio input is the feature people miss. Connect an MP3/WAV/OGG track to audio and it guides the generated video's sound - voice, music, whatever the track carries. Note the interaction with generate_audio: the driving track is ignored when generate_audio is false, so if you've wired audio and hear nothing, that's the reason. The other quirk is video-edit billing: when you supply a video, the edit inherits the clip's length (length_seconds and aspect_ratio are ignored) and you're billed the input duration at the full per-second rate on top of the output - a 10-second clip costs 20 seconds' worth.

The inputs that matter

prompt - required; the edit instruction when video is connected (e.g. "change this to an 8-bit pixel style").
video - the clip to edit (2–15s, ≤100MB). Switches the node to edit mode; can't combine with image/end_image.
image / end_image - start and optional end frame for image-to-video.
reference_images - up to 4, free, used only with video to guide the edit (person or style to apply).
generate_audio - defaults true, no price impact.
audio - the driving-audio track; ignored unless generate_audio is on.
length_seconds - 5/10/15; ignored in video-edit mode.
resolution - 720p or 1080p.

Output is video (VIDEO) - 2025+ ComfyUI for the core video type.

How it works

Standard pack mechanics: media serialized and uploaded, a job created, polled until done, MP4 downloaded into the temp folder. Video jobs take minutes; the progress bar is your only signal.

Install and key setup

ComfyUI Manager → BudgetPixel, or:

cd ComfyUI/custom_nodes
git clone https://github.com/budgetpixel-ai/comfyui-budgetpixel.git
pip install -r comfyui-budgetpixel/requirements.txt

Only requests. API key from a paid plan via BUDGETPIXEL_API_KEY or budgetpixel.json; keys never in workflow JSON.

Where people get burned

The three traps: grabbing the image node instead of this video node (same display name), expecting the driving audio to work with generate_audio off, and the full-rate double-billing on video edits. Cost-estimate before an edit run, force_rerun to defeat ComfyUI's cache, and remember interrupted jobs still bill server-side.

CategoryBudgetPixel/Video

Inputs (11)

Name	Type	Default	Description
prompt	STRING		Text description of the video, or the edit instruction when `video` is supplied (for example "change the video to an 8-bit pixel style").
aspect_ratioopt	COMBO	16:9	Aspect ratio. Ignored when a start image or an input `video` already sets the frame.
audioopt	AUDIO		Optional driving-audio track (MP3/WAV/OGG) for the generated video. Provide a public audio URL or an uploaded-file URL from POST /v1/uploads (passed by URL, not inlined). Ignored when `generate_audio` is false.
end_imageopt	IMAGE		Optional end frame, used together with `image` (the start frame) to interpolate the video between the two frames. Same input forms as `image`.
generate_audioopt	BOOLEAN	true	Generate audio with the video (default true; no price impact).
imageopt	IMAGE		Optional start frame for image-to-video. Provide a public image URL, a data URI, raw base64, or an uploaded-file URL from POST /v1/uploads. Omit for text-to-video. Can't be combined with `video`.
length_secondsopt	COMBO	5	Video length in seconds. Ignored for video editing, which inherits the input clip's length.
reference_imagesopt	IMAGE		Optional reference images (up to 4, free) used only with `video` — they guide the edit, for example supplying the person or style to apply to the clip. Each item is a public image URL, a data URI, raw base64, or an uploaded-file URL from POST /v1/uploads.
resolutionopt	COMBO	720p	Output resolution. Pricing varies by resolution — see `resolution_pricing` in GET /v1/models.
videoopt	VIDEO		Optional input video to edit (2-15s, MP4/MOV, ≤100MB). Supplying it switches the request to video editing: the model applies `prompt` to this clip. Provide a public video URL or an uploaded-file URL from POST /v1/uploads (videos are passed by URL, not inlined). Billed on the INPUT duration at the full per-second rate in addition to the output — estimate via POST /v1/cost. Can't be combined with `image`/`end_image`.
force_rerunopt	BOOLEAN	false	—

Outputs (1)

Name	Type	Description
video	VIDEO	—