News

Google Launches Gemini 4 Neon for Paying Users at Twice Hydrogen’s Price

Google released Gemini 4 Neon on November 18 for Google AI Pro and Ultra subscribers and the paid Gemini API only, at $3 per million input tokens and…

4 min read
Google Launches Gemini 4 Neon for Paying Users at Twice Hydrogen’s Price

On 18 November 2026, Google DeepMind announced Gemini 4 Neon in a post by Koray Kavukcuoglu on the DeepMind blog, introducing a new model built for complex software engineering and scientific workflows. Priced at double the cost of Hydrogen and restricted exclusively to paying tiers, Neon expands Google’s frontier capabilities while leaving free users completely on the sidelines. In an increasingly crowded lineup, it also creates fresh confusion over which model paying customers actually need.

Where Neon Sits in a Crowded Lineup

Neon is Google’s fourth Gemini 4 model to ship in roughly four weeks. As noted by The Decoder, while Carbon and Argon target lighter rivals, Neon represents Google’s top-tier, Opus-class answer. Arriving right after our coverage of Gemini 4 Hydrogen, releasing four models in a single month makes choosing needlessly difficult. Operating under model code gemini-4-neon, it offers a 1-million-token context window, a 64,000-token output limit scalable to 1 million via long decode, and multimodal inputs across text, images, audio, video, and PDFs. Output is strictly text-only, without image or audio generation. Supported tools include computer use, code execution, file search, and grounding via Google Search and Maps.

Sponsored

Tired of juggling ten tabs? ToolSuite bundles the AI workflow tools power users rely on — in one place.

Try ToolSuite Now

Access: Paying Customers Only

Rollout began on 18 November across select paid platforms. In the Gemini API, Neon is available in Google AI Studio exclusively for paid tiers, with the API free tier completely excluded. Paid Google AI Pro and Ultra subscribers can access it in Google Antigravity and from the model drop-down in the Gemini app. It also appears in Android Studio and Gemini for business, with a cyber defense edition for vetted defenders under the Fairwind Program coming shortly. Free Gemini app users receive nothing. Google omitted any app country list or timeline for Jules, the Gemini CLI, and Search AI Mode. Google’s pricing page says paid-tier API data is not used to improve its products.

Pricing: Twice Hydrogen, Below Upcoming List Rates

According to the Gemini API pricing page, Neon costs twice as much as Hydrogen, but will sit below Carbon and Argon once their introductory rates expire on 1 December.

ModelInput / 1M TokensOutput / 1M Tokens
Gemini 4 Neon$3.00$15.00
Gemini 4 Hydrogen$1.50$7.50
Carbon & Argon (Introductory)$2.00$10.00
Carbon & Argon (From 1 Dec)$4.00$20.00

Output pricing includes billed thinking tokens. Implicit context caching costs $0.15 per million tokens (a 95 percent discount), while batch requests run at half price ($1.50 input, $7.50 output, $0.075 cached). Search grounding grants 5,000 free monthly requests across Gemini 3+ models, then costs $14 per 1,000.

What Gemini 4 Neon Means for Arab Users

Google’s announcement named no languages and provided no Arabic evaluations. Google reports its lowest fabrication rate yet in internal Antigravity testing alongside strong multi-step instruction following, and cites external leaderboards for coding, agents, finance, and legal work. However, these claims come exclusively from English tasks and offer no Arabic measure.

The Gemini web app lists Arabic among supported languages, and the official available-regions page lists Egypt, Saudi Arabia, the UAE, Morocco, Iraq, Jordan, Lebanon, Algeria, Libya, Sudan, and Yemen, while Syria is not on the list. Because Google AI Pro and Ultra subscriptions are billed in local currencies at varying rates, readers should verify pricing on their local Google One page, as local fees were not checked.

We suggest readers run a four-part evaluation: test an Egyptian or Gulf dialect prompt; summarize an extensive Arabic PDF or contract leveraging the 1-million-token window; analyze an image containing Arabic typography; and compare identical outputs against Hydrogen. Because Hydrogen costs half as much, Neon must prove its worth in real Arabic workflows.

The Verdict: Who Should Switch and Who Should Wait

Neon is a powerful tool on paper, but its premium status rests on Google’s own claims until independent benchmarks appear. Developers running complex multi-step coding, autonomous agents or heavy document analysis should test Neon on the paid API against Hydrogen on their own tasks. The 1M-token window alone is no reason to switch: Hydrogen has one the same size. However, for everyday conversational tasks, standard app development, and routine summarization, Gemini 4 Hydrogen remains the practical champion at half the cost. Unless your workload genuinely demands Opus-class reasoning, staying on Hydrogen is the smarter financial choice.

Categories: News
Share:

Comments

0

No comments yet.