Skip to main content

Image & Video Generation

Unlocks creative production at scale

Maggie avatar
Written by Maggie
Updated today

Overview

Now users can generate outstanding still and video creative directly within Triple Whale. Bringing creative production together with creative analysis promises to simplify workflows and help brands get more value from their data.


Key Benefits

  1. Easily generate new photos and videos that are inspired by your winning ads

  2. Go from creative idea to prototype in mere minutes

  3. Transform any product photo into vivid video ready for Youtube

Get Connected

  1. In chat

    1. Go to triplewhale.com/chat and describe what you want Moby to generate

    2. Make sure to specify whether you want a still photo or video

    3. When requesting video, also specify how long you want the video to be

  2. In agents

    1. Go to triplewhale.com/workflows/create and setup any Get Data, Analyze Data, or Report steps needed for your use case

    2. Add an Action step and select “Generate Video”

    3. Type in positive and negative prompts, and specify how long you want the video to be

Best Practices

  1. Pair image and video generation with computer vision in order to build on prior experience

  2. Tell Moby to review the image or video assets he created for you, provide constructive criticism, iterate based on that feedback, and keep going in that cycle until he is satisfied

  3. All assets that Moby generates can be downloaded as well as accessed centrally by clicking the plus button in Moby chat

  4. Refer to our Prompting Guide for further insights about how to best prompt Moby.


Video Prompt Framework

While the SMART framework keeps Triple Whale prompts business-focused, Google’s new Veo-3 video model offers a complementary way to think about creative prompts to generate beautiful video content.

Video Prompt Framework (using Veo3 for rich media & beyond)

Building-block

What it covers

Mini-example

1. Subject

Who/what is on screen (people, animals, objects, combos).

“A playful golden-retriever puppy”

2. Action

The movement or behaviour.

“Leaps to catch a frisbee”

3. Scene / Context

Location, time of day, weather, era.

“At sunset on a misty beach”

4. Camera Angle

Eye-level, low/high angle, bird’s-eye, etc.

“Low-angle tracking shot”

5. Lighting & Style

Natural vs. artificial light, overall tone/mood, artistic style or color palette.

“Golden-hour glow, cinematic 35 mm look”

6. Ambiance

Atmospheric & textural details (fog, rain, grain, neon, etc.).

“Soft fog rolls across wet cobblestones”

7. Temporal Elements

Pacing (slow-mo, time-lapse), subtle evolution.

“Time-lapse day-to-night skyline”

8. Audio (opt.)

Dialogue, ambient noise, SFX.

“Waves crashing, distant seagulls”

9. Post-Prompt Helpers

Gemini as “expert prompter” or “second-pair-of-eyes”.

Ask Gemini to refine or QC the prompt.

Quick Rule of Thumb

Think of SMART as the business objective and Veo-3’s nine elements as the creative blueprint. Use them together when you need both strategic insight and vivid storytelling (e.g., campaign videos, ad mock-ups, product-storyboards).

Best Practices using Video (Veo3)

  1. Use cinematic language – terms like match-cut, montage, or split-diopter give the model clearer direction.

  2. Be laser-specific, avoid filler words. Clear, direct phrasing reduces ambiguity.

  3. Dialogue tip: Use a colon (:) after the speaker to prevent on-screen subtitles.

  4. Generate multiple aspect ratios (16:9, 9:16, 1:1) for omnichannel performance.

  5. One scene per prompt – break multi-step stories into separate clips for sharper results.

How this applies inside Triple Whale

  • Creative brief → Agent prompt: Use the Veo-3 table to structure any prompt that needs rich descriptive context (e.g., “Write ad-copy and suggest matching video concepts”).

  • SMART overlay: Once you’ve drafted the creative prompt, run it through SMART to ensure it’s measurable and time-bound before handing it to an Agent.

  • Iterate with Moby: Treat Moby as your in-house prompt QA—ask it to tighten language, flag missing data context, or suggest alternate styles.

Frequently Asked Questions

1. Which models does Moby use to generate videos?

  • Veo3 for video and Nano Banana for images (both from Google).

2. Is Moby good at preserving the details of product photography?

  • Yes

Did this answer your question?