← Google Gemini
Creator Magic
I Let Google Omni AI Edit My Videos
⏱ 45 min video · 3 min read1 Jul 2026Worth watching
TL;DR
Creator Mike Russell live-tests Google Gemini Omni Flash Preview, a new video generation and editing AI model that dropped less than 24 hours before the stream. He demos real-time video transformations including turning himself into a cheesy British radio DJ, generating fake product infomercials, and using drawn arrows to direct motion graphics — all from still images or short clips.
Key points
1
Gemini Omni Flash Preview hit general availability during the stream and is accessible via Google AI Studio (aistudio.google.com) or programmatically via the Gemini API with Claude Code
2
The model can transform a still image of a person into a video with consistent likeness, voice transformation, and character changes — such as turning the creator into a British radio DJ with a sung jingle
3
You can draw arrows or markings on an image and Omni will generate motion that follows those directional cues, enabling creative control over generated video movement
4
Gemini Omni Flash is a thinking model with adjustable thinking levels (high/low), making it capable of reasoning through complex multi-step video generation prompts
5
The cost for an entire live session of heavy video generation was approximately $1 USD initially, with a total daily API spend of around $26.50 before hitting a hard quota limit
Actionable insights
→
Use Google AI Studio at aistudio.google.com to access Gemini Omni Flash Preview immediately — drag and drop an image, add a prompt, and generate video without any coding
→
Add 'same face, same likeness' to your prompts to improve consistency of the subject across generated video frames
→
Draw directional arrows or markings on your source image before uploading to give Omni spatial instructions for motion direction in the output video
→
Use Claude Code with the Gemini API docs loaded into agents.md for programmatic batch video generation when AI Studio becomes unstable or rate-limited
→
Set the thinking level to 'high' in Gemini Omni Flash for complex prompts involving character transformation, voice, and multi-stage sequences
Notable quotes
“If I had access to a tool like this when I was working in radio, I probably wouldn't have shown up to the radio station. I'd have just made things like this.”
“I am not going to sleep tonight. I might just spend the rest of the live stream generating cheesy British radio DJs.”
“You can give direction — you can actually take something and draw arrows and waves for the sea and it will generate really cool motion graphics that follows exactly the arrow that I drew on the screen.”
Worth watching?
✅
Worth watching the full video?
Watch if you want to see live unscripted demos of what Gemini Omni Flash can actually do with real images — the key capabilities and costs are all captured here, so skip it if you just need the practical setup steps.
Saved you some time? The creator still deserves a like.
Want this for your own channels?
Add the channels you follow. Every new video summarised and in your inbox the moment it drops. From £4/month.
Try it free