Google Introduces Gemini 4 Neon: Three Times Faster Than Carbon

· AI · Cem Koyluoglu

While its competitors dominated the AI world all summer, Google, which had been silent for a few months, has now launched a counterattack. Less than a month ago, the company released its new flagship ship

TL;DR Google launches Gemini 4 Neon, a mid‑tier AI model that claims faster, cheaper performance on complex coding and reasoning tasks compared to its flagship Gemini 4 Carbon. Neon supports multimodal inputs and a massive one‑million‑token context window, but its real‑world performance remains unverified. Pricing is higher than Carbon, though promotional discounts make Neon more cost‑effective for certain workloads.

Key Highlights • Gemini 4 Neon targets complex coding, analytical reasoning, and multi‑step workflows, positioning itself between the flagship Carbon and budget Hydrogen. • Google claims Neon runs at 60 % of Carbon’s cost and three times its speed, and even outperforms Carbon on certain tasks in internal tests. • The model is multimodal—handling text, image, audio, and video—and supports a context window of up to one million tokens. • Pricing is $3 per 1 million input tokens and $15 per 1 million output tokens; Carbon is discounted to half price until November 30, making Neon cheaper to run during the promotion. • Neon is available to Gemini app subscribers on Google AI Pro or Ultra, via a paid API, and through Google Antigravity and corporate solutions.


Google Unveils Gemini 4 Neon, Expanding Its AI Portfolio

After a quiet period, Google has entered the AI race with a new flagship model, Gemini 4 Neon. The company had previously launched Gemini 4 Carbon and the budget‑friendly Gemini 4 Hydrogen, and now adds Neon to cover the mid‑tier market.

Targeted Use Cases and Performance Claims

Neon is positioned for “complex coding tasks, analytical reasoning, and multi‑step workflows.” Although Carbon remains the most powerful model Google has released to date, the company reports that Neon outperforms Carbon on certain tasks. Google also claims that Neon completes tasks at 60 % of Carbon’s cost and at three times the speed. These figures are based on internal tests, and the model’s performance in external settings has not yet been verified.

Benchmark Scores

Neon achieved a score of 61 on the Artificial Analysis Intelligence Index and 65.4 on CursorBench 4.1. Google states that the model is “competitive with leading models in its class.” The AI supports text, image, audio, and video inputs, making it a multimodal system with a context window of up to one million tokens.

Pricing Structure

Pricing for the Gemini 4 Neon API is set at $3 per 1 million input tokens and $15 per 1 million output tokens. These rates are higher than the current promotional prices for Carbon, which are $2 and $10 respectively. However, Google will offer Carbon at half price until November 30, making the more powerful Neon cheaper to run during the promotion period. The company’s claims that Neon is more cost‑effective than Carbon for task completion may not be fully reflected in the raw numbers.

Availability

Neon is available to users of the Gemini application who subscribe to Google AI Pro or Ultra. Developers can access the model through a paid API, and it is also offered via Google Antigravity and Google’s corporate solutions.

Source: https://www.neowin.net/news/google-launches-gemini-4-neon-claims-its-three-times-faster-than-carbon/

← All tech news