Skip to main content

Gemini 4 Carbon

Google 🧠 Language Model

Google's most capable reasoning model, for long agentic coding, science and math: a 1M-token context, text, image, video and audio in, text out, with native tool use and three thinking levels.

Specifications

Input modalitiestext, image, video, audio
Output modalitiestext
Capabilitiestools, reasoning, coder
Relative speed●●○○○ Relaxed
Context window1,048,576 tokens
Max output65,536 tokens
Input price$8.00 / 1M
Output price$40.00 / 1M

Pricing

$8.00 / 1M input tokens  ·  $40.00 / 1M output tokens

Platform credit pricing — live rates on the Models page.