ai/list101

Gemini 4 Carbon

aistudio.google.com

A multimodal reasoning model for sustained coding agents, knowledge work and scientific applications through Google's paid developer platform.

Llm ModelsDeveloper Tools
Visit website
Pricing
Paid
Available on
API, Google AI Studio
Category
Llm Models

What is Gemini 4 Carbon?

Gemini 4 Carbon is a reasoning model available through the paid Gemini API and Google AI Studio. Google positions it for sustained software engineering, multimodal knowledge work, scientific reasoning and cybersecurity defense.

The stable model ID is gemini-4-carbon. It accepts text, images, video, audio and PDF inputs and produces text, allowing developers to combine several source formats within one reasoning workflow.

Key features

Adjustable thinking levels

Choose low, medium or high reasoning effort to balance token use and latency. Medium is the recommended default.

Large shared context

Input and output share a 1,048,576-token context window.

Long decode

Background execution can generate beyond the default synchronous output limit, using the remaining shared context budget.

Agent tools and structured output

Supports function calling, structured output, code execution, computer use and file search.

Search and Maps grounding

Supports grounding with Google Search and Google Maps, with tool pricing separate from model token usage.

Who is it for?

Software engineers investigating complex changes

Evaluate an agent workflow for analyzing code, planning revisions and checking proposed implementations.

Knowledge workers comparing source material

Ask questions across supplied documents, images and recordings.

Researchers exploring scientific questions

Compare reasoning approaches and request structured explanations of supplied material.

How to use Gemini 4 Carbon

  1. Open Google AI Studio or prepare access to the paid Gemini API.
  2. Select gemini-4-carbon as the model.
  3. Start with medium thinking and provide the task and relevant source material.
  4. Set an output budget that leaves room for thinking and the final answer.
  5. Evaluate the response and compare thinking levels for the intended workload.

Gemini 4 Carbon pricing

Paid

Usage-based token billing with no Carbon free tier. Standard introductory rates apply through November 30, 2026 and increase on December 1. Batch, Flex, Priority, caching and tools have separate rates.

Standard paid tier through November 30, 2026

USD 2 per 1 million input tokens; USD 10 per 1 million output tokens
Metered usage, not a monthly subscription

Output pricing includes thinking tokens. Applies to Standard inference through November 30, 2026; caching and tool charges are separate.

Standard paid tier from December 1, 2026

USD 4 per 1 million input tokens; USD 20 per 1 million output tokens
Metered usage, not a monthly subscription

Output pricing includes thinking tokens. Applies to Standard inference starting December 1, 2026; caching and tool charges are separate.

Check current pricing ↗

What to know before you use it

  • Carbon has no API free tier. Thinking tokens and internal tool instructions used with grounding are billable, and tools can add charges.
  • Default synchronous output is limited to 65,536 tokens. Longer output requires long decode and remains constrained by the shared input-and-output context.
  • Image generation, audio generation and the Live API are unsupported. Output is text.
  • Context caching is implicit only; explicit caching is unsupported.

Frequently asked questions

Which thinking level should I start with?

Google recommends medium for most tasks. Evaluate all three levels and choose the lowest effort that meets the workload's quality needs.

Does the maximum output allowance exclude thinking?

No. The configured output budget covers thinking and the final answer together, so reasoning can consume space otherwise available for the response.

Can Carbon use Batch, Flex or Priority inference?

Yes. The model supports all three options, with separate pricing from Standard inference.