Generate Image and Video

Describe what you want, your Chief of Staff makes it. Ad creative, thumbnails, social posts, product shots, short reels — anywhere you'd otherwise hire a designer or video editor.

You don't pick a model. You don't fiddle with settings. You just describe it.

Images cost 10 credits each. Videos start at 250 credits (longer, higher-resolution or premium-model videos cost more) — and your CoS shows a confirmation card with the exact price, model, resolution, length and shape before it runs one. That price is what you're charged, and that is the video you get, so you're never surprised.

What it does for you

Images on demand:

• Photorealistic shots, illustrations, logos, simple diagrams
• Square for feed posts, landscape for thumbnails, portrait for stories
• Attach a photo and ask for a version of it — "put me in front of this background", "make this warmer" — and the result keeps what's in your photo. If your photo can't be used, nothing is made and you aren't charged
• Generated in seconds and shown right in chat
• Ready to download, drop into a post, or hand to your team

Short videos for ads and social:

• Reels and TikToks at 9:16, YouTube at 16:9, square for feed
• A few seconds up to fifteen — enough for an ad hook or a product demo
• Comes with ambient sound and effects by default — background noise, not a spoken voiceover. Want it silent? Just say so ("make it silent", "no audio") and it renders without sound
• Works for product demos, brand b-roll, ad creative, explainer clips
• What you approve is what renders. If the video model on your card can't take the request, nothing is made and you aren't charged — your CoS tells you why and can offer another model at its own price and resolution, on a new card you approve first. Your video is never quietly moved to a different model or a lower resolution
• If a finished video comes back in a different shape from the one on your card, you aren't charged for it, and your CoS tells you its real shape
• If your prompt is longer than the video model accepts, your CoS shortens it with you before anything renders instead of cutting off the end
• Renders typically take 3-6 minutes. A fast one appears right in chat; a slower one keeps rendering in the background and shows up in the same chat message and your Asset Library when it finishes — you don't have to keep the tab open, and you're only charged when it succeeds

Three ways to drive a video:

• Just type a prompt and go
• Hand it a photo to use as the first frame so a still image animates to life. A video started from a photo takes that photo's shape, so a portrait photo makes a portrait video. If your photo's shape doesn't match the shape you asked for, your CoS tells you before the confirmation card and offers to render in the photo's shape or to use a photo in the shape you want. A different-shaped last-frame photo gets cropped to match the first
• Hand it a reference image so the video matches your brand look

On-brand without thinking:

• Pick a vibe: cinematic, social ad, explainer, product demo
• Your CoS knows your brand — colors, voice, audience — from your Brain
• Re-roll, refine the prompt, or swap the style without starting over
• Edit the prompt and run it again until it looks right

Voiceovers and narration (needs ElevenLabs):

• A generated video's audio is ambient sound and effects. It does not read a script aloud, so it is never a narration track
• To get a spoken voiceover, connect ElevenLabs in Integrations. Your CoS will then generate the audio for you — ask it to "narrate this" or "make an mp3 of this script" and it picks a voice and produces the file
• Ask for a narrated video and your CoS will tell you which half it can do: it makes the visuals, ElevenLabs makes the spoken track, and you lay one over the other
• ElevenLabs is connected with an API key you create in ElevenLabs (Profile → API Keys). Keys start with sk_ and are only shown once, when you create or rotate them — copy the key itself, not the key's name or ID

Saved automatically:

• Every image and video lands in your Asset Library
• Tagged with the prompt that made it so you can find it later
• Linked to the conversation that produced it
• Download or share with one tap

How you use it

1. Open the chat.
2. Describe what you want — "a photo of a coffee cup on a wooden desk, morning light" or "a 9:16 ad showing a product unbox in 8 seconds."
3. Your CoS picks the right tool and generates it. For a video, approve the confirmation card first (it shows the exact price). Images take seconds; videos typically take 3-6 minutes and keep rendering in the background if they run long — the credits are only charged when the render succeeds.
4. The result shows up inline in chat. Don't love it? Ask for a tweak — "make it warmer" or "switch the background to a kitchen."
5. When it's right, grab it from chat or open it from your Asset Library.

For videos: be specific. Subject, action, setting, mood, camera movement. The more you say, the better the result.

If you only do one thing today: open the chat and ask your CoS for "a 9:16 social ad showing a happy customer using my product, 8 seconds, social ad style." Watch what comes back.