WordOnThisWordOnThisSearch
Submit
Gemini 4 Carbon logo

Gemini 4 Carbon

Multimodal Google model for complex software engineering and mathematical reasoning.

Visit website

By Google DeepMind

Gemini 4 Carbon official website preview
Loading website screenshot…The official website preview is loading.

What is Gemini 4 Carbon?

Carbon is a Google model built for complex software engineering and mathematical reasoning. It accepts text, images, audio, video, and PDF inputs within a shared 1,048,576-token context window, supports function calling, and generates text output up to a 65,536-token default synchronous limit.

What can Gemini 4 Carbon do?

Multimodal Input and Text Output

Accepts text, images, audio, video, and PDF inputs in a shared 1,048,576-token context window and generates text up to a 65,536-token default synchronous output cap.

Function Calling Support

Supports function calling to connect external tools during software engineering and mathematical reasoning tasks.

Who is Gemini 4 Carbon for?

  • Complex Software EngineeringAnalyze large codebases, technical documentation, and diagrams using the shared 1,048,576-token context window and function calling.
  • Mathematical ReasoningSolve multi-step mathematical problems from text or PDF inputs with detailed text output.

How much is Gemini 4 Carbon?

View official pricing ↗
paid

Gemini API standard paid tier: USD $2 per million uncached input tokens and $10 per million output tokens through November 30, 2026. From December 1, 2026, the rates are USD $4 per million uncached input tokens and $20 per million output tokens. This model has no free API tier. Output pricing includes thinking tokens.

Frequently asked questions

What is the Gemini API model ID?

Use gemini-4-carbon to call this model.

What inputs and context window does it support?

The model accepts text, images, audio, video and PDFs in a shared 1,048,576-token context window. It produces text, with a default synchronous output limit of 65,536 tokens.

How did it perform on the Coding Agent Index?

Artificial Analysis reported a Coding Agent Index score of 55 with Antigravity CLI at high effort. The score applies to that model, effort and harness combination. Source: https://artificialanalysis.ai/articles/gemini-4-neon-intelligence-gain/

What people are saying

Public experiences and discussions, with links to the originals.

No discussions collected yet.

Explore 11 products that share at least one second-level category.

View all 11 alternatives

Help keep this listing accurate

Found something incorrect? Suggest a correction for our editors. It’s free; no account or email is needed.