Google Releases Gemini 0.36 To General Public
Google's Gemini 4 Carbon is now available to everyone, offering advanced features at competitive prices. The model's pricing is set to increase on December 1st.
Original reporting · simonwillison.net
Original: “llm-gemini 0.36”
Read “llm-gemini 0.36” at Simon Willison →
How this was written. GDS wrote this article with a language model from Simon Willison’s reporting (model: @cf/meta/llama-3.3-70b-instruct-fp8-fast); it describes that reporting rather than adding to it. The original, linked above, is the full story.
Key points from Simon Willison’s report
- Google's Gemini 4 Carbon is now available to the general public, offering advanced features at competitive prices.
- The model's pricing is set to increase on December 1st, reaching $4 and $20 per million inputs and outputs, respectively.
- Gemini 4 Carbon has been demonstrated to be capable of generating complex animations, 3D models, and interactive content, including open-world 3D games.
- The model's capabilities have been recognized by prominent figures, including Sundar Pichai.
- Potential users should evaluate the model's value proposition before the price increase takes effect.
Google has announced the general availability of Gemini 4 Carbon, a model previously limited to a subset of cyber defenders. According to Simon Willison, this release marks a significant expansion of access to the technology. Gemini 4 Carbon is Google's best iteration of the Gemini model to date, with competitive pricing at launch.
The pricing for Gemini 4 Carbon is currently set at $2 per million inputs and $10 per million outputs, comparable to other models like GPT-6.1 Sol and Claude Sonnet 5.5. However, these prices are scheduled to double on December 1st, reaching $4 and $20 per million inputs and outputs, respectively, to match the pricing of Opus 5.5.
Simon Willison reported that Gemini 4 Carbon offers three levels of thinking effort: low, medium, and high. The model has been tested using Google's Antigravity IDE, demonstrating its capabilities in generating complex animations and 3D models. For example, it was used to create a three.js animation of a mech standing up and moving around Manhattan, Queens, and Brooklyn, utilizing OpenStreetMap data.
The Antigravity IDE test also involved prompting the model to build a standalone HTML file, which was then screen-captured as a video. This showcases the model's ability to produce interactive and engaging content. Furthermore, the model's capabilities extend to building open-world 3D games, as demonstrated by its use in creating a game where players explore Google's Mountain View campus as a pelican riding a GBike.
The game was built using the /swarm feature in Google Antigravity, which coordinated subagents to model 3D assets in Blender, build the browser game in Three.js with WebGL shaders, and pair procedural Web Audio sound effects with Lyria music. This example highlights the potential of Gemini 4 Carbon in generating complex and interactive content.
According to the source, the capabilities of Gemini 4 Carbon have been recognized by prominent figures, including Sundar Pichai, who tweeted about the model's achievements. The release of Gemini 4 Carbon to the general public may have significant implications for the development of AI-powered content creation and cyber defense applications.
The future development and adoption of Gemini 4 Carbon will likely depend on its performance and the response of users to its capabilities and pricing. As the model's pricing is set to increase, it may be important for potential users to evaluate its value proposition before the price change takes effect.
Questions and answers
What is Gemini 4 Carbon?
Gemini 4 Carbon is Google's latest iteration of the Gemini model, offering advanced features at competitive prices.
How much does Gemini 4 Carbon cost?
The current pricing is $2 per million inputs and $10 per million outputs, but it is scheduled to double on December 1st.
What are the capabilities of Gemini 4 Carbon?
Gemini 4 Carbon can generate complex animations, 3D models, and interactive content, including open-world 3D games, as demonstrated by its use in Google's Antigravity IDE.

