Google Veo 3 in Google Ads: Generate Video Ads From Product Images at No Cost
Before May 6, 2026, using Veo 3 meant going through the API or Vertex AI and paying roughly $0.40 per second for Veo 3 Fast or $0.75 per second for the full model. That was fine for teams with engineering resources and a production budget. For most DTC brands running Google Ads without a video team, it was not a realistic option.
Google changed that by folding Veo 3 image-to-video directly into Google Ads Asset Studio. Every active account gets it at no extra cost. You upload product images, write an optional scene prompt, and get back a roughly 10-second HD video with auto-generated ambient audio. The video lands in your Asset Library and is immediately usable in Performance Max, Demand Gen, and YouTube campaigns.
Where to Find It in the Interface
The entry point is: Google Ads dashboard, Tools menu, Asset Library, Create Asset, Video, then “Generate with Veo 3.” If the account has been active (at least some spend in the past 30 days), the option should be there without any additional setup.
Two reasons you might not see it. First, Google is still rolling out to all accounts and some are receiving access later than others. By end of May 2026, access should be universal. Second, dormant accounts that have not run ads for several months can fall into an inactive status where some newer features are not shown. Running a small campaign for a few days usually resolves that.
The input options are straightforward: upload up to three product images in JPEG or PNG format, and an optional text prompt to direct the scene style. The image itself drives the visual output; the prompt steers things like setting, lighting, and motion feel. Images at 1024px or higher produce noticeably cleaner results.
Step-by-Step: From Product Image to Ready-to-Run Asset
The full workflow from upload to campaign-ready video takes about five minutes of active time, plus a 90 to 120 second generation wait.
Step one, go to Asset Library inside Google Ads, click Create Asset, select Video, and choose “Generate with Veo 3.”
Step two, upload your product images. You can upload up to three, and each one generates a separate video. Keep backgrounds clean or solid-colored if possible. Veo 3 handles compositing better when the product boundary is clear.
Step three, write a scene prompt if you want to direct the output. A useful format is: setting, lighting, motion descriptor. For example, a seller of outdoor drinkware might write “Morning light, camping scene, condensation slowly forming on the exterior.” Leaving the prompt blank is allowed, but the resulting scenes will be more generic.
Step four, click Generate and wait roughly 90 to 120 seconds. The preview will appear in the interface. If the result does not work, regenerate. There is no per-generation limit disclosed by Google, so iterate freely.
Step five, if the video is acceptable, save it to Asset Library. It is then available to add to any compatible campaign type.
| Parameter | Spec |
|---|---|
| Input formats | JPEG or PNG, up to 3 images |
| Output length | Approximately 10 seconds |
| Output resolution | HD (exact pixel dimensions not published) |
| Audio | Auto-generated ambient sound, cannot be replaced |
| Cost | Included with active Google Ads account, no extra charge |
| Compatible campaign types | Performance Max, Demand Gen, YouTube In-stream |
What It Produces Well and Where It Falls Short
The strongest use case is product-centered showcase clips. A product sitting on a surface in a relevant setting, with natural lighting variation and gentle movement, is exactly what Veo 3 handles reliably. Outdoor gear in an outdoor scene, home goods under warm interior lighting, tech accessories on a desk: all of these produce usable output consistently. The motion is subtle rather than cinematic, which actually suits most product ad contexts where the product needs to stay legible throughout.
White or solid-color product backgrounds make a meaningful difference. Veo 3 does the scene compositing by separating subject from background, and a clean background gives it cleaner edges to work with. Complex or patterned backgrounds tend to produce artifacts at the product boundary.
Three things it does not do well. Human faces: any attempt to generate creative with people in it will produce uncanny or distorted results. Google does not recommend it for that use case, and the output confirms why. Talking head formats, testimonials, and spokesperson-style ads are not the right application. On-screen text: Veo 3 is inconsistent when text needs to appear in the video itself. Titles, CTAs, and brand names should be added in post using a separate video editor. The 10-second fixed length: if your campaign format requires 6-second bumpers or 15-second cuts, there is currently no output length control. What you get is approximately 10 seconds, and you trim from there if needed.
How It Compares to Standalone Tools
For teams already paying for Runway, Kling, Pika, or similar tools specifically for Google Ads video, here is a direct comparison.
| Dimension | Google Ads built-in Veo 3 | Runway, Kling, Pika, etc. |
|---|---|---|
| Cost | Included with Ads account | $20-$100/month, or per-credit |
| Video length | Fixed ~10 sec | Typically 4-16 sec, adjustable |
| Camera control | Basic prompt direction only | Keyframing, camera motion, reference image support |
| Audio | Auto ambient, not replaceable | Most support voiceover or music upload |
| Human generation | Not recommended | Some have dedicated realistic-human models |
| Workflow | Direct to Asset Library, one step | Download then re-upload to Ads |
If your main need is product-focused video for Google Ads campaigns, the built-in tool is sufficient and saves you a workflow step. If you need more precise creative control, multi-platform output (TikTok and Meta in addition to Google), or human-led video formats, a standalone tool is still warranted. Most advertisers end up using both for different parts of the creative mix.
Before You Push It Live
A few checks worth running before adding a generated video to a live campaign.
The first two seconds have to do something. YouTube skippable ads give viewers the option to skip after five seconds. Demand Gen has similar patterns. If the opening shot is a slow dissolve from a static product image, viewers are gone. Check that there is movement or a clear visual hook in the opening frame.
The product needs to stay visible throughout. Veo 3 sometimes moves the subject toward the edge of the frame during the scene generation, or applies heavy depth-of-field blur late in the clip. Scrub through the full video to confirm the product remains prominent and legible from start to finish.
Add brand elements in post. Nothing in the generated video will carry your brand name, CTA text, or any overlay. Export it, add those in a basic video editor, and then import the branded version to Asset Library. Sending the raw Veo 3 output to a campaign without brand identification misses the entire point of running video ads.
For YouTube In-stream, a video around 10 seconds puts you at the lower boundary for non-skippable formats. For YouTube Shorts and vertical inventory, note that Veo 3 currently outputs landscape-oriented video. A vertical crop will work if the product is centered, but check that the reframing does not cut off key product detail.
Source: Google Ads — Create video assets with Veo.
Related Articles
Google AI Brief in Practice: Three Layers of Natural Language Control for AI Max
AI Brief lets advertisers steer Gemini inside AI Max using plain English across three layers: Messaging Guidelines, Matching Guidelines, and Audience Guidelines. Here is how to use each layer and what cross-border sellers should write in each field.
Kling AI 3.0 for Cross-Border Ad Video: Multilingual Lip-Sync and Brand Consistency
Kling AI 3.0 launched February 5, 2026 with two features that matter most for cross-border sellers: reference-based product consistency across shots, and native lip-sync across English, Chinese, Japanese, Korean, and Spanish. Here is how to put both to work.
Google Retiring DSA in September 2026: Ecommerce Seller Migration Guide to AI Max
Starting September 2026, Google officially retires Dynamic Search Ads and auto-migrates all eligible campaigns to AI Max for Search. AI Max requires no keywords — AI automatically matches search intent. What cross-border ecommerce sellers need to prepare before the deadline.