NEWS · MODELS · #3774
Google showcases five community projects built with Gemini 4 Carbon
Google highlighted five developer projects using Gemini 4 Carbon, including a CUDA/C++ meadow with 52,000 flowers, a WebGPU real-time simulation of 1,048,576 particles, a fragile stained-glass block-stacking simulation, 14 interactive tools generated from messy public records, and a 3D reconstruction of a 19th‑century optics lab with 92 instruments. Google says developers can access Carbon via the Gemini API paid tier, Google AI Studio, Android Studio, Google Antigravity, the Gemini app, AI Mode in Search, and CodeMender, with availability varying by subscription (e.g., AI Ultra).
KEY POINTS
- Google highlighted five developer projects using Gemini 4 Carbon, including a CUDA/C++ meadow with 52,000 flowers, a WebGPU real-time simulation of 1,048,576 particles, a fragile stained-glass block-stacking simulation, 14 interactive tools generated from messy public records, and a 3D reconstruction of a 19th‑century optics lab with 92 instruments.
- Google says developers can access Carbon via the Gemini API paid tier, Google AI Studio, Android Studio, Google Antigravity, the Gemini app, AI Mode in Search, and CodeMender, with availability varying by subscription (e.g., AI Ultra).
- The examples illustrate Gemini 4 Carbon's ability to handle large-scale simulation, complex scene generation, and data‑cleaning workflows, and show integration across Google's developer tools and subscription tiers.
WHY IT MATTERS
The examples illustrate Gemini 4 Carbon's ability to handle large-scale simulation, complex scene generation, and data‑cleaning workflows, and show integration across Google's developer tools and subscription tiers.
SOURCES & TIMELINE
2These ambitious real-world projects simulate live physics, untangle messy data, turn a single reference image into a 3D design, and more. Gemini 4 Carbon was built for tasks that push the limits of a model’s intelligence, and builders are putting it to the test. Since we rolled out Carbon more broadly last week, we've seen the community lean into the model’s ability to explore thoroughly, reason deeply, and run ext…
As part of a new direction for Stitch, we’re introducing a new experiment, Stitch Loop: a long-running product agent that connects to your codebase and live metrics to continuously improve what you’ve shipped. Coding agents let you build and ship new features fast, but their job ends the moment your pull request merges. Once code hits production, the real work starts: patching vulnerabilities, speeding up checkout, …
Google released Gemini 4 Neon, promising "frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows." The model is live today through the API at $3 per million input tokens and $15 per million output tokens. That puts it above Carbon and Argon, which currently share an introductory rate of $2/$10 per million tokens through November 30 b…