Google Showcases Four Gemini 3.8 Flash Projects Built by Its Community
The projects range from an animated painting to an interactive transmission. They show what builders made, not whether the technical models are accurate.
Loading page…
The projects range from an animated painting to an interactive transmission. They show what builders made, not whether the technical models are accurate.
Listen to this story
Google is using community-built demos to make Gemini 3.8 Flash’s claims about multi-step reasoning and coding easier to inspect, rather than announcing a new model release. The projects include visualizations and interactive 3D builds; notably, a transmission prototype offers 10 camera views and four display modes. Gemini 3.8 Flash launched September 2, and Google says it can take extra reasoning steps and call tools repeatedly on complex tasks. The showcase demonstrates what builders can produce, but does not show
Ashutosh Shrivastava paired Gemini with Google Antigravity to map live satellite, space-station and orbital-rocket paths.
Noctus animated Seigaiha waves; Emily used four prompting phases to build a T. rex skeleton and request accuracy checks.
Hakm’s transmission prototype has 10 camera views and four display modes, plus clickable parts and an explanatory side panel.
Google has published four community projects made with Gemini 3.8 Flash, turning its claims about the model’s reasoning and coding abilities into examples readers can inspect. The builds span live orbital maps, animated art, a dinosaur skeleton and an interactive engineering model.
Ashutosh Shrivastava used the model with Google Antigravity to map the live paths of satellites, space stations and orbital rockets. Noctus used its ability to work with different kinds of input to turn Seigaiha waves, a traditional Japanese pattern, into an animation resembling an ink painting.
The other two builders pushed toward models a user can inspect. Emily prompted Gemini through four phases to make a T. rex skeleton, setting requirements, ruling out shortcuts and asking for accuracy checks. Hakm built an automatic-transmission prototype with clickable parts and a side panel explaining them.
Gemini 3.8 Flash is not a new release this week: Google introduced it on September 2. Google says it improves on 3.7 Flash in software engineering, tasks involving AI agents and reasoning through several steps. For complex work, the company says, it can take extra reasoning steps and call tools repeatedly. That helps explain why Google chose projects requiring more than a single written answer for this showcase.
The showcase demonstrates what these builders created, but it does not validate the scientific or mechanical accuracy of the results. That distinction matters most for the skeleton and transmission, where an engaging model and a correct one are not necessarily the same thing.
Google says builders can try Gemini 3.8 Flash in Google AI Studio or its other integrated tools. The practical test is whether a project stays useful when someone checks its underlying details, not just whether it looks finished on screen.
Loading discussion...
Join the conversation
What would make it convincing to you?
Be the first to share a perspective or an experience.
Reader comments
Newest comments first. Replies stay oldest first.