Which tool to generate UGC-style video ads for products at scale?

No single tool fits every UGC-style ad pipeline. The right one depends on what you start from (a script, a product page link, product photos, or a brief), who appears on camera (a stock AI actor, your own avatar, or footage a video model generates), how many languages the ads need, and whether you want an app with templates or an API your own code calls once per product. At catalog scale, also check for a batch or queue endpoint, a price unit you can multiply by your volume, and a spend limit you control.

The table below compares six tools from their own pages, read on 2026-09-29. Sume facts come from the Format catalog, Create a run, Bulk runs, and Generate avatar video docs pages.

Which tools make UGC-style product ads, and how do they compare?

Rows are alphabetical, not ranked. Each vendor cell comes from that vendor's own pricing page, API docs, help center, or home page, with the price unit as the page states it. How HeyGen and Synthesia differ from Sume in more depth is in Sume vs HeyGen and Sume vs Synthesia.

Which one should I pick?

Match the tool to the input you already have, the presenter you want, and the way you want to pay:

• Arcads if you want an AI actor who holds your product or shows your app, and you are fine buying monthly credits; its API works through products, scripts, and a generate call.
• Creatify if your starting point is a product page link or a single product image; its API prices per 30 seconds or per image in credits.
• HeyGen if the same ad must go out in many languages, or you want to submit up to 100 videos in one batch call paid from prepaid dollars.
• Synthesia if you want a script-led avatar video with 160+ languages and voices, and the Creator plan's yearly API minutes cover your volume.
• Runway if you want to generate the footage with a video model and cut the ad yourself, billed in per-second credits.
• Sume if you want a ready-made UGC recipe fed with a brief and product images, with a spend cap per run and a server-side queue; an avatar holding your product reading your script; or Kling and Seedance models on one API.
• Sume is not the right pick if the presenter must speak a language other than English on Avatar 1.0 (current code speaks English only), or if you need an avatar video above 720p, the documented resolution.

Do I need an ad app or an API pipeline?

An ad app gives you templates and an editor, which suits a person reviewing each ad. When a catalog feed produces the inputs, an API lets your code start one job per product and variant, cap the spend, and collect results by webhook.

On Sume, the API path has three levels. A catalog Format takes a brief and product images and leaves the model choice to the Format's tools. Avatar 1.0 takes your exact script, an avatar, and a product image. POST /v1/videos calls a video model directly, including kling-3, seedance-2.5, seedance-2-mini, seedance-2, seedance-2-fast; frame_images with first_frame turns a product photo into the opening frame on models that list it. GET /v1/videos/models returns each model's durations, aspect ratios, supported frame images, and pricing_skus, as described in Video generation.

Which Sume catalog Formats make a UGC-style ad?

Sume's catalog answers at the reserved sume handle. Two of its slugs end in ugc, and a third names a product usage demo. A Format is a saved recipe the agent applies, so read a Format's description with GET /v1/formats/sume/{slug} before you call it; a slug not in the catalog answers 404 format_not_found. The whole catalog is in Ready-made Formats for product video.

Call one with a key that has formats:write. Put the brief in instruction and product shots in attachments, up to 30 images the agent can see; the Format's tools choose the image, video, and audio models. Send an Idempotency-Key on every create, and add communication.webhook_url to receive one signed format.run.terminal POST when the run completes or fails.

curl -sS -X POST "https://api.sume.com/v1/formats/sume/sume-close-camera-ugc/runs" \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: sku-4411-ugc-v1" \
  -d '{
    "instruction": "Short vertical ad for the attached product.",
    "attachments": [
      { "type": "input_image", "image_url": "https://example.com/product.jpg" }
    ],
    "generation_spend_cap_usd": 20
  }'

How do I make one with an avatar, a script, and a product image?

Use Avatar 1.0 when you want to choose the presenter and write every word. POST /v1/avatar-1.0/talking-video names a ready avatar with avatar_handle and exactly one of script or video_inputs; add product_image for the product, or omit it for a productless video. scene can be { "type": "prompt" } with a prompt or { "type": "photo" } with an image_url, and media fields must be fetchable public HTTPS URLs. No avatar yet? See how to create a reusable AI avatar.

The docs' multi-scene example is shaped like a UGC ad: a spoken hook, a four-second silence beat named demo, and a spoken call to action, each with the background prompt "Casual bedroom framing, native UGC lighting". How to write that kind of script is in UGC ad scripts for AI avatar videos. To review first-frame stills before paying for a full render, use avatar video previews.

curl -X POST https://api.sume.com/v1/avatar-1.0/talking-video \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: ugc-avatar-sku-4411-v1" \
  -d '{
    "avatar_handle": "acme_host",
    "script": "I kept this in my bag all week. Here is why.",
    "product_image": "https://example.com/product.png",
    "scene": { "type": "prompt", "prompt": "Casual bedroom framing, native UGC lighting" },
    "quality": "plus",
    "aspect_ratio": "9:16"
  }'

How do I make many variants per product on Sume?

Testing UGC ads means many versions of each product ad: different opening hooks, presenters, and calls to action. On Sume each version is its own job or run:

• Hooks and scripts: each script is one Avatar 1.0 job, estimated at 4–60 seconds. Ways to vary the opening are in Hook variations for UGC ads.
• Presenters: each avatar video uses one resolved avatar, so another presenter means another job with another avatar_handle.
• Catalog Formats for many products: POST /v1/formats/{handle}/{slug}/bulk-runs queues 1–100 items, each the same body as a single run with its own attachments and generation_spend_cap_usd, and keeps concurrency (1–16) of them in flight; how to collect the results is in Sume Format bulk runs.
• Capacity: workspace generation concurrency (1 on Free, 4 on Pro, 8 on Startup, 20 on Scale in the docs' plan table; read generation_limits.concurrency_limit) applies to every child run and avatar job. Jobs past it can be accepted as queued while queue capacity remains; a full queue answers 429 queue_full.

curl -sS -X POST "https://api.sume.com/v1/formats/sume/sume-close-camera-ugc/bulk-runs" \
  -H "Authorization: Bearer $SUME_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: ugc-batch-2026-09-29" \
  -d '{
    "concurrency": 4,
    "items": [
      {
        "instruction": "Short vertical ad for the attached product.",
        "attachments": [{ "type": "input_image", "image_url": "https://example.com/sku-1.jpg" }],
        "generation_spend_cap_usd": 20
      },
      {
        "instruction": "Short vertical ad for the attached product.",
        "attachments": [{ "type": "input_image", "image_url": "https://example.com/sku-2.jpg" }],
        "generation_spend_cap_usd": 20
      }
    ]
  }'

Can the ads speak other languages?

Not on Avatar 1.0: in current code it speaks English only. For another language, generate the voiceover with TTS 1.0 (POST /v1/tts-1.0/generate) and set language for every non-English transcript, then make a talking clip from a still and that audio with VEED Fabric 1.0 (POST /v1/veed/fabric-1.0). Fabric's audio must be Sume-hosted and under 10 MB, so pass the TTS result's mirrored audio file. The API pricing rate card lists TTS at $0.0475 per 1,000 characters and Fabric at $0.1875 per audio second (720p). The fields are in the API reference, and the steps in Dub an AI avatar video with lip sync.

What does a UGC ad cost on Sume?

On the Format path, generation that counts against the run's cap is metered at the API pricing rates, and generation_spend_cap_usd bounds it: up to $500, null runs at $500, and 0 is rejected. The receipt's usage.billable_amount_usd_micros excludes the agent's own LLM turn, so it is not the run's total cost. An avatar video is priced per second by quality and by whether you send a product image. With a product image the rate is $0.194 per second for standard, $0.258 per second for plus, and $0.58 per second for max.

What are the limits?

These bounds come from the same docs pages and code:

• Avatar scripts and multi-scene plans must land at an estimated 4–60 seconds. Shorten a longer script or split it into several jobs.
• Current avatar execution supports one resolved avatar per final video and expects scene backgrounds to resolve to one shared scene.
• Avatar aspect_ratio is 1:1, 3:4, 9:16, 4:3, or 16:9; resolution is currently 720p.
• A generated clip shows a model's rendering of the product. If the ad's claim is how the real product looks in use, such as fit, application, or a before-and-after, film the real product for that part.