Sales enablement video: short talk-track clips made with AI

A sales enablement video is a short internal video that teaches sales reps something they use on calls: the pitch, the answer to a common objection, or what changed in the latest release. Make one clip per talk track, keep each under a minute with one presenter, and re-render a clip when its message goes stale.

An AI avatar turns each talk track into a video from its script, so updating the pitch means editing text, not booking a shoot. On Sume each clip is a talking-video job, with multi-scene video_inputs when you want a hook, a pause, and a close. Facts come from Generate avatar video, read on 2026-09-28, plus Sume's current code where noted.

Which sales enablement videos should a team make first?

Start with the moments reps ask about most. One clip each:

• The short pitch: what the product does and for whom, in a few sentences.
• Objection answers: one clip per objection, such as price, timing, or an incumbent tool.
• Discovery questions: the questions to ask on a first call.
• Release briefings: what's new this month and how to position it.

How do I structure one clip?

Hook, one point, close. A plain script covers most talk tracks. When you want a pause before the answer, ordered video_inputs scenes, spoken and silent, build one video whose total planned duration must still land in the 4-60 second window; the multi-scene avatar video API covers the scene fields. Give every scene the same background: current execution expects scene backgrounds to resolve to one shared scene.

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: enablement-price-objection-v1" \
  -d '{
    "avatar_handle": "enablement_coach",
    "aspect_ratio": "16:9",
    "video_inputs": [
      { "id": "hook",
        "voice": { "type": "text", "script": "When a buyer says the price is too high, start here.", "duration": 4 },
        "background": { "type": "prompt", "prompt": "Quiet meeting room" } },
      { "id": "pause",
        "voice": { "type": "silence", "duration": 2 },
        "background": { "type": "prompt", "prompt": "Quiet meeting room" } },
      { "id": "close",
        "voice": { "type": "text", "script": "Ask what the problem costs them today, then compare.", "duration": 5 },
        "background": { "type": "prompt", "prompt": "Quiet meeting room" } }
    ]
  }'

Can the product appear in the video?

Yes, as a reference image. product_image takes a public HTTPS image URL; omit it for a productless video. Clips with a product image bill at a separate per-second rate, shown in the table below. AI avatar video with your product and background covers what to expect from the product in frame, and adding B-roll to an avatar video covers cutting in other footage.

How do I keep the clip library current?

• Keep one avatar handle so every clip has the same presenter.
• When a talk track changes, edit its script and render that clip again. Use a new Idempotency-Key: the docs say to reuse a key only for the same operation and payload.
• Store each clip's media.sume.com URL next to its talk track, and swap it when you re-render.
• Check the framing of a new setup with an avatar video preview, which returns first-frame stills before the full render.

What are the limits, and what does it cost?

• In current code the avatar route speaks English only.
• resolution is currently 720p, with one presenter per video.
• No product screen recordings: there is no public upload route for local files, so a software demo needs another tool.