Video Creation lets you turn a text idea, a single image, or two key frames into a finished short-form video — perfect for ads, social reels, and product showcases.
There are three video modes in MorningAI: Text to Video, Frames to Video, and Frame-to-Frame (which is actually Frames to Video with an End Frame uploaded — same surface, more powerful flow). All three live in the same studio.
The 30-Second Tour
Open the Studio, choose Create a video, and you'll land in a single creator with three modes you can switch between at the top. Pick a mode, write your prompt, set your aspect ratio and duration, click Create. Most videos finish in around two minutes.
Outputs land in your gallery and your File System automatically. You don't have to download to keep them.
Mode 1 — Text to Video
Type a description of what you want. Optionally pull in a product, a style, or a persona to shape the scene. Click Create.
Best when you don't have visuals yet — concepting, mood pieces, abstract launches, anything where you're starting from a sentence.
One thing to know: you can't attach products in Text mode. The hint in the UI says: "Select Frames or Ingredients mode to use products." If a product is the subject of your video, switch modes.
Duration in Text to Video: typically 5 or 10 seconds. Audio: available.
Mode 2 — Frames to Video (image-to-video)
Upload a Start Frame — a real photo, a generated image, or a product packshot from your library — describe the motion you want, and click Create.
Best for animating a hero photo, making a packshot move, or bringing a static ad to life.
Frame-to-Frame: add an End Frame for smooth transitions
Upload both a Start Frame and an End Frame. The platform animates a smooth transition between them. Best for product transformations (closed → open), before/after demos, scene transitions, and morph effects.
Two practical tips for Frame-to-Frame: keep the two frames consistent (same camera angle, same lighting, same product placement), and use the prompt to describe the motion, not just the scene. "Camera slowly orbits the product, light moves left to right, steam rises from the cup" beats "a coffee cup."
Accepted file types for frames: JPEG, PNG, WebP. Duration in Frames to Video: 4, 6, or 8 seconds. If you've added reference images elsewhere in the brief, the End Frame may be disabled with the tooltip "Not compatible with reference images."
Mode 3 — Ingredients to Video
A fourth mode in the dropdown: Ingredients to Video uses up to 3 reference images for style or element guidance. Locked to 8 seconds and 16:9 aspect ratio. Useful when you want the model to take cues from existing imagery without animating it directly.
The Settings That Change Everything
The gear icon opens Settings:
- Aspect Ratio — Landscape (16:9) or Portrait (9:16). Get this right at generation time; re-cropping later loses quality.
- Duration — depends on mode. Text: 5 or 10s. Frames: 4, 6, or 8s. Ingredients: locked to 8s.
- Resolution — when available.
- Outputs per prompt — 1 to 4. Generate multiples; the first take is rarely the keeper.
- Audio — toggle sound on/off. Audio + duration are linked; turning audio on may force a different duration.
Match Aspect Ratio to Where the Video Will Live
- 9:16 → Reels, TikTok, Shorts, Stories.
- 1:1 or 4:5 → Instagram feed.
- 16:9 → YouTube, LinkedIn, Facebook video, landing page hero.
Pick aspect ratio at generation time, not in post. Re-cropping a 16:9 to 9:16 chops the frame; generating natively at 9:16 gives the model the right canvas to compose for.
What Happens After You Click Create
A progress bar runs through Initializing → Processing → Generating → Almost ready → Finalizing. Most videos finish in around two minutes.
Outputs appear as cards in the output panel, ready to:
- Play and review.
- Publish (kicks into the same Publish Flow as social posts).
- Download as MP4.
- Favorite (saves to your favorites for fast retrieval).
- Trash or restore.
- Upscale to 4K — separate generation, only do this on the keeper.
Upscale to 4K — Only the Winner
Generate at standard resolution, browse the takes, pick the best, then click Upscale to push it to 4K for hero placements. Upscale is a separate generation, not a free post-process — treat it as a second step.
Pro Tips
- Prompt for motion, not just for the scene. The model needs verbs.
- Pull from Product DNA in Frames mode — your real packshots become animated product videos in seconds.
- Generate 2–4 outputs per prompt. More variants, fewer credits burned re-rolling.
- Sound off ≠ no music. Many advertisers prefer silent generations because they layer their own audio in editing — and silent generations are faster.
- Switching modes resets some inputs. Set your mode first, then build the prompt.
Limits and Gotchas
- Requires an active subscription or available credits.
- Videos take noticeably longer than images — plan for ~2 minutes per generation.
- You can't use Text-to-Video to feature a product. If a product is the subject, use Frames or Ingredients mode.
- Audio + duration are linked. Pick audio first, then duration.
- Upscale is a separate generation. Treat it like a second step, not a free post-process.
- Credits ledger labels your generations as "Create Video without Sound" or "Create Video with Sound."
FAQ
How long does video generation take?
Most generations finish in around two minutes. Audio-on tends to take longer. Upscale to 4K is its own generation on top of that.
Why can't I attach a product in Text mode?
Text-to-Video doesn't accept reference images. To feature a real product, switch to Frames mode and upload a packshot, or use Ingredients mode for style guidance.
Why is duration locked to 8 seconds?
Ingredients to Video is fixed at 8s and 16:9 by the model. Frames to Video gives you 4, 6, or 8s; Text to Video gives 5 or 10s. Switching modes is the way to unlock different durations.
What happens if I run out of credits mid-generation?
In-progress generations finish. You'll be prompted to buy credits or activate a subscription before starting the next one.
