In brief
- Gemini 4 Neon is beginning to roll out through the Gemini API and Google AI Studio.
- The model has a 1 million-token context window; API pricing is $3 per million input tokens and $15 per million output tokens.
- Google showed the model creating a photorealistic 3D laboratory workbench in Google Antigravity using Blender MCP.
- Gemini 4 Neon is available in the Gemini app to Google AI Pro and Ultra subscribers.
Google DeepMind announced Gemini 4 Neon on 18 November 2026. The model is beginning to roll out across various Google products, including the Gemini API and Google AI Studio. The company positions Neon for tasks that require multiple steps, such as complex software development, knowledge work and scientific research.
Context window and pricing
Gemini 4 Neon has a 1 million-token context window. Google says the model can evaluate different types of content—including spreadsheets, PDFs, charts, audio and video—within the same context. API pricing is $3 per million input tokens and $15 per million output tokens. Cached input tokens are priced at $0.15 per million.
In evaluations shared by Google, Neon scores 61 at the xhigh reasoning level on the Artificial Analysis Intelligence Index, ranking third overall. The company also reports that the model ranks first on the Artificial Analysis Coding Agent Index, with a score of 58.6. These are the measurements reported in Google's announcement; actual performance may vary depending on the task and toolchain.
Google also says that, at the default high reasoning level, Neon costs 32% less per task than Carbon. This percentage refers to the task cost calculated in the company's own comparison, rather than the model's token pricing.
Which products is it available in?
Gemini 4 Neon is available to developers through the paid tier of the Gemini API and in Google AI Studio. Enterprise users can access the model through Gemini for business, via the API or model selection. A Google AI Pro or Ultra subscription is required to try Neon in Google Antigravity; the model option in the Gemini app is also available to Pro and Ultra subscribers. Cybersecurity use will be made available in the coming days to eligible defenders in the Fairwind Program.
The company also highlights features such as sharing progress during long sessions, reducing unnecessary file reads and managing context. The announcement includes internal test and benchmark results for using these capabilities in production workflows; as no independent verification is provided, the results should be regarded as Google's claims.
Impact on architecture, interior design and archviz workflows
The announcement's standout example for 3D production shows Gemini 4 Neon using Blender MCP in Google Antigravity to create a photorealistic laboratory workbench that visualizes model names. This demonstrates that the model can participate in a 3D production task through a tool; it does not, on its own, mean that this is a built-in Blender feature or a generally available integration.
For architecture firms and visualization teams, a potential use case could be carrying out multi-step tasks involving 3D production tools through an agent. However, the example is presented in the context of Google Antigravity and Blender MCP. The announcement does not provide details on compatibility with other software, hardware requirements or how these capabilities could be set up in different workflows. It is therefore important to verify access requirements and compatibility with the tools in use before deploying it in existing projects.
Sources
1 sourceSource texts are not republished; short quotes are marked, everything else is our own summary and commentary.
What stands out about Gemini 4 Neon for architecture and visualization teams is less its ability to respond to a single prompt and more its claim that it can carry out multi-step tasks alongside tools. The Blender MCP example makes this potential tangible, but it does not mean the model will work out of the box and without issues in every 3D application.
For offices in Turkey, decisions should take into account the per-task cost of API usage, access routes that require a Pro or Ultra subscription, and compatibility with existing tools. The announced prices are per token; actual total costs depend on workflow usage. A cautious approach is to run a small-scale trial before purchasing or incorporating it into production.
Frequently asked questions
How much does the Gemini 4 Neon API cost?
According to Google's announcement, input tokens cost $3 per million and output tokens cost $15 per million. Cached input tokens cost $0.15 per million.
Can Gemini 4 Neon be used in Blender?
Google shared an example of the model creating a 3D laboratory workbench using Blender MCP in Google Antigravity. The announcement does not say this is a built-in Blender integration available to everyone.
Who can access Gemini 4 Neon?
The model is available through the paid tier of the Gemini API and Google AI Studio. Google AI Pro and Ultra subscribers can use Neon in Antigravity and the Gemini app.



