Vidu Q4 Preview 30% off until Nov 30: plan a test batch and a fallback

Vidu takes 30% off Q4 Preview image-to-video and reference-to-video through Nov 30. Budget a test batch and keep a Sume fallback for the same first frames.

4 min readSume
All posts

Vidu is taking 30% off image-to-video and reference-to-video generation with Q4 Preview through November 30, according to the launch coverage of the GlobeNewswire release. The discount names those two modes, not text-to-video. If you want to test Q4 Preview, run a short batch inside that window; if you also want a fallback on Sume, you can generate the same shots there from the same first frames.

What the window covers

The release states the discount runs through November 30 for image-to-video and reference-to-video. It does not say, on the page we read, whether the discount applies to every resolution or to the $0.014 launch rate. Treat the exact post-discount price as unconfirmed until you see it in Vidu's console.

Discount facts, read 2026-10-08
ItemWhat the source saysConfidence
Discount size30%Stated in the GlobeNewswire release
Modes coveredimage-to-video, reference-to-videoStated in the GlobeNewswire release
End dateNovember 30Stated in the GlobeNewswire release
Applies to 4K at $0.39Not statedCheck Vidu console
Sume discountNone documentedSume rates are list x 1.25

A test batch with a Sume control

The cheapest way to learn whether Q4 Preview is worth switching to is a controlled batch: the same ten first frames and the same prompts on both sides.

Steps:

  • Pick ten first frames and one prompt for each.
  • Run them on Vidu Q4 Preview through its image-to-video endpoint before November 30.
  • Run the same frames on Sume as frame_images with frame_type: "first_frame", for example on wan-3.0 at 720p, which is $0.125 a second.
  • Compare the outputs by eye, then compare the cost per accepted clip.
  • Keep an Idempotency-Key on each Sume request, so a retry returns the original job instead of a second charge.
curl -X POST https://api.sume.com/v1/videos \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Idempotency-Key: q4-control-001" \
  -H "Content-Type: application/json" \
  -d '{"model":"wan-3.0","prompt":"Slow push-in, steam rising from the cup","resolution":"720p","duration":6,"frame_images":[{"type":"image_url","image_url":{"url":"https://example.com/cup.png"},"frame_type":"first_frame"}]}'

Scoring the batch

Decide the scoring before you spend. Count a clip as accepted when it is usable without a retry, and divide total spend by accepted clips. A cheaper model that needs three tries per usable clip can cost more than a dearer one that lands first time.

Record the prompt, the first frame, the model, the resolution and the cost for every run in one sheet. When the discount ends, you can then re-price the same accepted set at the new rate and see whether the choice still holds.

Keep the sample honest by mixing easy and hard shots. Include at least one close-up face, one fast action beat and one shot with text or a logo in the frame, because those are where models differ most. Ten easy landscapes will make every model look good and tell you nothing about the money you will spend later.

What Sume does not do

Sume does not run a promotion that matches Vidu's, and it cannot call Vidu for you. After November 30 the comparison should be redone at full price. A six-second Wan 3.0 clip at 720p is $0.75 on Sume, which gives you a fixed number to compare against Vidu's post-promotion quote.

Sources

Related posts

More in Pricing

All Pricing posts

Written by Sume