AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: A Lower Bill, A Small Gain: Assessing Google’s Gemini 4 Neon on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get the latest gadgets delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

Google released Gemini 4 Neon on November 18. Artificial Analysis rates it one point above Gemini 4 Carbon at its highest effort setting, while reporting lower cost and shorter task times; the index still places two other models ahead. Results vary by effort setting and benchmark, and Google’s comparison table does not include those index leaders.

Google released Gemini 4 Neon on November 18, presenting it as a model that balances performance and efficiency. In Artificial Analysis’s independent results, Neon scored 61 on its Intelligence Index at the xhigh setting, one point above Gemini 4 Carbon at high, while an index task cost less and took less time. The index still ranks GPT-6.2 Astra and Claude Fable 5.2 ahead of Neon.

Artificial Analysis reports that an index task with Neon at xhigh costs $6.44 and takes 21.5 minutes, compared with $7.76 and 37.1 minutes for Carbon at high, using standard list prices. At Neon’s default high setting, it scores 60, matching Carbon’s reported high-setting score, at $3.74 per task. The figures suggest that the choice of effort setting can alter both cost and performance as much as the model change itself.

Google lists Neon at $3 per million input tokens and $15 per million output tokens, below Carbon’s standard rates of $4 and $20. Carbon is discounted to $2 and $10 through November 30, according to the source report, so the list-price comparison does not describe the temporary offer. Artificial Analysis also reports that Neon generates output at about 110 tokens per second, versus about 55 for Carbon, but uses roughly 15% more output tokens per task, reducing some of the price benefit.

On the Artificial Analysis Coding Agent Index, Neon scores 58.6, close to Claude Fable 5.2 at 58.5 and Claude Opus 5.5 at 58.2. Its reported run costs $6.86 per task, against $15.47 for Fable 5.2 and $12.61 for Opus 5.5. Results are not uniform across tests: Neon’s xhigh score on Terminal-Bench 4.0 is 63.6%, below Carbon’s 65.2% at high. Google’s own evaluation table marks Neon best in 16 of 19 rows, but compares it with Claude Opus 5.5 and GPT-6.2 Sol, not the two models ranked above Neon by Artificial Analysis.

At a glance
reportWhen: Released November 18, 2026; figures rev…
The developmentGoogle released Gemini 4 Neon, a model adapted from Gemini 4 Carbon that shows a small independent benchmark gain alongside lower listed prices and shorter task times.

Gemini 4 Neon

Four models in seven weeks

Google’s Gemini 4 releases side by side, each at its best tested setting in the independent Artificial Analysis index. The best value in each row is outlined.

ArgonCarbonHydro.Neon ReleasedSep 30Oct 28Nov 12Nov 18 Index50605461 Cost/task$4.97$7.76$3.16$6.44 Min./task29.737.115.321.5 Output $$10$20$7.50$15 Carbon $2/$10 discount ends Nov 30Neon: +1 point, $1.32 less per task
  1. Four models in seven weeksArgon on September 30, Carbon on October 28, Hydrogen on November 12, Neon on November 18. Neon is adapted from Carbon.
  2. A small step in the indexNeon at xhigh scores 61, one point above Carbon at high. It ranks third among model families, behind GPT-6.2 Astra and Claude Fable 5.2.
  3. A bigger step in cost and timeOne index task costs $6.44 instead of $7.76 and takes 21.5 instead of 37.1 minutes. Neon’s tokens are 25% cheaper than Carbon’s standard price.
  4. What to watchUntil November 30, Carbon still sells at a discounted $2/$10. Google’s own table compares Neon with Claude Opus 5.5 and GPT-6.2 Sol, not with the index leaders.

Source: Artificial Analysis, Gemini 4 Neon model page and article, accessed Nov 19, 2026 (Intelligence Index v5.0.1, cost and time per task at standard prices, release dates); Google, “Introducing Gemini 4 Neon”, Nov 18, 2026; Gemini 4 Neon model card. Output price in US$ per million tokens, standard rates. Redrawn by Thorsten Meyer AI.

Graphic: Powered by Thorsten Meyer AI

Lower Costs, With Setting Trade-Offs

For developers and businesses choosing a model for repeated coding or agent tasks, cost and completion time can matter as much as a small benchmark-score difference. Artificial Analysis’s figures indicate that Neon may deliver a similar index score to Carbon at high for less money and less time, although its xhigh result costs substantially more than its high result. Buyers will need to test their own workloads rather than assume that the highest setting is automatically worthwhile.

The near tie on the Coding Agent Index, paired with lower reported run costs than the two Claude models in that comparison, makes Neon relevant to teams weighing model performance against usage expense. But benchmark results do not establish that it will be cheaper or better for every application. Token use, task length, reliability requirements and the price actually paid—including any temporary Carbon discount—can change the calculation.

Products Worth Considering

Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Bluetooth 5.3, USB-C Charging, and IPX4 Water Resistance

Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Bluetooth 5.3, USB-C Charging, and IPX4 Water Resistance

DJI Mini 4K Drone Combo with 4K Camera, 3-Axis Gimbal, and Two Batteries

DJI Mini 4K Drone Combo with 4K Camera, 3-Axis Gimbal, and Two Batteries

As an affiliate, we earn on qualifying purchases.

Neon Builds on Gemini 4 Carbon

Neon is Google’s fourth Gemini 4 model in about seven weeks, following Argon on September 30, Carbon on October 28 and Hydrogen on November 12. Google describes Carbon as suited to difficult, long-running projects and Hydrogen to fast everyday execution, positioning Neon between them. Its model card says Neon is adapted from Gemini 4 Carbon and directs readers to Carbon’s documentation for details about architecture, training data and hardware.

Google lists support for text, images, audio and video input, text output and a one-million-token context window. The default maximum output is 64,000 tokens, with up to one million available through background long decode. The model has four effort settings—low, medium, high and xhigh—with high as the default; its stated knowledge cutoff is March 2026. Google says it is available through the Gemini API in AI Studio and selected business and subscriber products. A cyber-defense version is being made available to trusted defenders in Google’s Fairwind Program, with eligible members expected to receive access over the following days.

Artificial Analysis characterizes the release as a small Intelligence gain over Carbon alongside lower cost per task. That assessment is based on its own benchmark and cost methodology, not a claim that Neon is broadly superior across all tasks. The source report also discloses that it was researched and drafted with AI assistance, including Claude, and says the article was not reviewed or sponsored by Google, Anthropic, OpenAI or Artificial Analysis.

“While Carbon tackles your hardest, long-horizon projects and Hydrogen handles fast everyday execution, Neon balances depth and efficiency.”

— Google’s launch post

Benchmark Coverage and Real-World Costs

It is not yet clear how Neon’s reported advantages translate to particular customer workloads. Artificial Analysis’s figures use its own tasks and methodology, and scores from different benchmarks are not interchangeable. Its index places GPT-6.2 Astra and Claude Fable 5.2 ahead of Neon, while Google’s published table omits both models from its comparison set. That difference in coverage limits direct reconciliation between Google’s table and the independent ranking.

Google reports that Neon has its lowest fabrication rate yet in internal Antigravity testing, but the source material says that measurement is not independently checkable and is distinct from Artificial Analysis’s hallucination test. On AA-Omniscience, Neon has a reported hallucination rate of 14.4%, compared with Carbon’s 11.4%; those figures should not be treated as equivalent to Google’s internal result. The source report also gives no full basis for extrapolating benchmark task costs to all users, whose bills will depend on usage and applicable pricing.

Testing Neon Against Workloads

The next useful evidence will come from independent testing across more tasks and from customer use after the November 18 release. Teams considering Neon can compare its high and xhigh settings on representative work, tracking completion quality, latency, token consumption and total cost rather than relying on a single index score. They should also account for Carbon’s stated discount through November 30 when comparing current prices.

Google says eligible members of the Fairwind Program will receive the cyber-defense version over the next few days. The source material does not specify further release dates or a broader access plan for that version. More detailed comparisons that include the models leading Artificial Analysis’s index would also help clarify how Neon fares against the strongest models in that ranking.

Key Questions

How much better did Gemini 4 Neon score than Carbon?

Artificial Analysis reports a score of 61 for Neon at xhigh, one point above Carbon at high, which scored 60. Neon at its default high setting also scored 60, matching Carbon’s reported high-setting score.

Is Neon cheaper to run than Carbon?

At standard list prices, Artificial Analysis estimates an index task costs $6.44 with Neon at xhigh and $7.76 with Carbon at high. Neon at high costs $3.74 per task. Carbon has a temporary listed discount through November 30, so current comparisons may differ.

Does Neon lead the independent model rankings?

No. Artificial Analysis ranks GPT-6.2 Astra and Claude Fable 5.2 ahead of Neon on its Intelligence Index. Neon is close to Claude models on the Coding Agent Index, but the rankings depend on the benchmark used.

Where can people use Gemini 4 Neon?

Google lists availability through the Gemini API in Google AI Studio and selected business and paid subscriber products, including Google Antigravity and the Gemini app for Pro and Ultra subscribers.

Source: ThorstenMeyerAI.com

You May Also Like

Exploring The Open Stack AI Power Behind A Simple Microduck Toy

Hugging Face unveils Microduck, an affordable, open, and forkable robot for embodied AI learning, marking a shift in robotics accessibility and security challenges.

An Inside Look At SpaceXAI’s Grok Bot And How To Experience It Early

SpaceXAI’s Grok chatbot is now in limited early beta, offering select users a first look at its conversational AI capabilities amid ongoing development.

South Korea’s Tax Exam Case: When AI Cheating Undermines Confidence In Credentials

A top tax accountant exam candidate confessed to using a camera and AI, prompting a police investigation and a review of national qualification exams.

Join The Badminton Community And Track Your Progress Easily

A mobile app for recreational badminton players to log matches, view rankings, and share highlights is being tested with local clubs, promising to unify scoring and social sharing.