Adjust short video clips with plain text requests
Gemini can change an uploaded clip one instruction at a time while trying to preserve the scene’s original look.
- What they did
- A creator tested Gemini Omni Flash in the Gemini app over a 36-hour stretch. They uploaded a clip, typed a single plain-English edit, and checked the result in the chat; later changes were layered onto earlier ones, and they could return to an earlier version if needed. They also looked at four public creator examples covering object removal, clothing replacement, a new background, and a flamingo-head change that carried across a meeting sequence.
- How it went
- The edits generally held onto the original lighting, camera position, and movement, and the flamingo-head change survived multiple people and cuts. Results were uneven, though: one creator judged their own demo as mediocre, and packing several edits into one request led to inconsistent output. Short clips, especially under 10 seconds, performed best, and pre-launch reporting said one generation used about 43% of a Pro user’s daily allowance, which worked out to around two completed rounds per day on the $19.99 plan.
- Tips
- For steadier results, ask for one change at a time and keep the clip brief. If a request could lead to deleting, sending, or submitting something, stop and get approval first.
Try it yourself with Gemini
I’m working with this video clip: [attach or describe clip]. Make only this change first: [describe one edit]. Try to preserve the scene’s lighting, camera position, and motion. After each turn, show the updated version and point out anything that looks weak or unnatural. If the next step would involve sending, deleting, or submitting anything, pause and ask me to approve it before continuing. Then wait for my next instruction: [second edit], [third edit].
Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.
Get 3 agent uses like this every morning
With the day's top AI news. Free, and one click to unsubscribe.
Related agent uses
Build a motion infographic from a blog post
A content creator used Gemini Canvas to convert a blog article into a branded animated infographic.
Build a Gemini CLI check for technical mistakes
A developer made a Gemini CLI command that scans blog drafts for technical errors and runs again after each revision.
Automate daily videos and form follow-ups in Make
A CTO used Make with Gemini to draft daily YouTube content, while another user routed Google Form leads into monday.com with automatic follow-ups.
Automate an AI newsletter from topic pick to inbox
An n8n schedule launches Gemini-written AI newsletters, cleans the markup, and emails the final version to subscribers.
Let an agent package one idea for multiple channels
A creator delegated script writing, thread drafting, and post repurposing to a CreaoAI agent.
Try an agent on scripts, long bios, and card mockups
A TechRadar writer compared Manus and ChatGPT on screenplay research, a long Adobe bio, and a business card.