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

🔍 Read the full analysis: How Gemini 4 Neon Balances Performance And Cost, According To Google 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, describing it as a model that balances depth and efficiency. Artificial Analysis measured a one-point score gain over Gemini 4 Carbon at the highest settings, alongside lower cost and shorter task times; results vary by effort setting and benchmark.

Google released Gemini 4 Neon on November 18, presenting it as a model that balances depth with efficiency. Independent benchmark data published by Artificial Analysis show Neon scoring 61 on the Intelligence Index at its xhigh setting, one point above Gemini 4 Carbon at high, while the measured task cost fell from $7.76 to $6.44 and average task time from 37.1 to 21.5 minutes.

The figures are not a like-for-like comparison at the same effort setting: Neon’s 61-point result is at xhigh, while Carbon’s 60-point result is at high. At high, the default setting for Neon, Artificial Analysis reports a score of 60 and a cost of $3.74 per index task. That matches Carbon’s high-setting score in the cited comparison, at a lower task cost. The index combines multiple evaluations; these figures are not a guarantee of performance on every workload.

Artificial Analysis attributes Neon’s cost and speed results to lower token prices and faster output, while also noting that Neon uses about 15% more output tokens per task than Carbon, offsetting some of the price difference. Neon’s standard prices are $3 per million input tokens and $15 per million output tokens, compared with Carbon’s standard $4 and $20. Carbon is discounted to $2 and $10 through November 30, according to the source material, so the standard-price advantage does not reflect that temporary promotion.

Google’s model card says Neon is adapted from Gemini 4 Carbon. It accepts text, images, audio and video, returns text, and has a one-million-token context window. Its four effort settings run from low to xhigh, with high as the default. Google lists access through its Gemini API, AI Studio and selected business and subscriber products. A cyber-defense version is being offered to eligible defenders through the company’s Fairwind Program.

At a glance
reportWhen: Released November 18, 2026; figures acc…
The developmentGoogle released Gemini 4 Neon, a model adapted from Carbon that the company positions as a faster, lower-cost option for completing complex tasks.

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

The Cost of More Capable AI

For teams running multi-step tasks, the reported combination of shorter completion times and lower per-task costs could affect which model they use and how often they use it. The relevant comparison is not simply a model’s benchmark score: total spending also depends on token consumption, effort settings, task length and the pricing available when a customer runs the model.

The coding results offer a more specific comparison. On Artificial Analysis’s Coding Agent Index, Neon scored 58.6, close to Claude Fable 5.2 at 58.5 and Claude Opus 5.5 at 58.2. The reported run cost for Neon was $6.86, versus $15.47 for Fable 5.2 and $12.61 for Opus 5.5. Those figures suggest a cost advantage in that test, but do not establish that Neon is better or cheaper for every coding task.

Google’s own evaluation table also needs to be read with its comparison set in mind. It marks Neon ahead in 16 of 19 rows against Claude Opus 5.5 and GPT-6.2 Sol. The table does not include Claude Fable 5.2 or GPT-6.2 Astra, which rank ahead of Neon on the cited independent index. Benchmark choices can shape how a launch appears, so readers should distinguish a vendor’s selected comparisons from broader third-party results.

Products Worth Considering

Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation

Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation

As an affiliate, we earn on qualifying purchases.

A Fourth Gemini Model in Seven Weeks

Neon is the fourth Gemini 4 model Google has released 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-horizon projects and Hydrogen as intended for fast everyday execution. It positions Neon between them, aiming to combine depth with efficiency.

Artificial Analysis’s summary characterizes the change from Carbon as a small Intelligence Index gain with lower cost per task. The model’s setting matters: at xhigh, Neon scored 61 for $6.44 per index task and took 21.5 minutes; at high, it scored 60 for $3.74 and took 14.8 minutes. Medium scored 55 at $1.90 and 8.3 minutes, while low scored 44 at $0.75 and 3.7 minutes. The figures were accessed November 19 and are stated at standard list prices.

Performance is not uniformly higher across tests. Neon at xhigh scored 63.6% on Terminal-Bench 4.0, below Carbon at high, which scored 65.2%. Artificial Analysis also reported a 14.4% hallucination rate for Neon on its AA-Omniscience test, compared with Carbon’s 11.4%. Google separately said Neon achieved its lowest fabrication rate yet in internal Google Antigravity testing; that is a company claim based on a different test that outside readers cannot independently verify from the supplied material.

““frontier-level intelligence with fast end-to-end task completion.””

— Google, in its Gemini 4 Neon launch post

Benchmark Gaps and Pricing Caveats

The available comparisons do not settle how Neon performs across all real-world uses. Artificial Analysis’s index and Google’s evaluation table use different tests and comparison sets; Google’s table omits Claude Fable 5.2 and GPT-6.2 Astra, which lead Neon in the cited independent rankings. Scores from different benchmarks should not be treated as interchangeable.

It is also unclear how representative the reported task costs and times are for a given customer. They depend on the test workload, model setting, token use and price assumptions. Carbon’s temporary discount through November 30 changes the immediate price comparison. The source material does not provide a detailed methodology for Google’s internal fabrication-rate result, nor does it establish how often Neon will outperform Carbon on users’ own tasks.

Testing Neon Across Real Workloads

Neon is available through the channels Google has listed, while eligible members of its Fairwind Program are due to receive access to the cyber-defense version over the following days, according to the source material. The immediate practical test for prospective users is to compare high against xhigh on their own tasks: the cited index data show a one-point score difference between those Neon settings, but a rise in task cost from $3.74 to $6.44.

Further independent evaluations could clarify how Neon compares with the models missing from Google’s table and whether its speed and cost advantages hold across different workloads. Readers should also check current prices directly: the reported Carbon discount ends November 30, and pricing or access terms may change. No additional benchmark release or later product milestone is specified in the source material.

Key Questions

What did Google announce?

Google released Gemini 4 Neon on November 18, 2026. It describes the model as balancing depth and efficiency and says it is adapted from Gemini 4 Carbon.

Does Neon score higher than Carbon?

On the cited Artificial Analysis Intelligence Index comparison, Neon at xhigh scored 61, one point above Carbon at high. At high, Neon scored 60. Because the 61-to-60 comparison uses different effort settings, it is not a fully matched comparison.

How much does Neon cost?

The reported standard rates are $3 per million input tokens and $15 per million output tokens. Artificial Analysis reported costs per index task ranging from $0.75 at low to $6.44 at xhigh. Actual spending depends on usage and setting.

Is Neon faster or better for coding?

Artificial Analysis measured shorter task times for Neon than Carbon in its index comparison. On its Coding Agent Index, Neon scored 58.6, close to Claude Fable 5.2 and Claude Opus 5.5. These results do not establish that Neon is faster or better for every coding workload.

What should readers watch for next?

Further independent tests may clarify how Neon compares with models not included in Google’s published table and how its cost and performance hold up across real workloads. Google’s reported Carbon discount is scheduled to end November 30, according to the source material.

Source: ThorstenMeyerAI.com

You May Also Like

Square Enix Surges In Global Coverage

Search interest in Square Enix has surged significantly, with 24 mentions in recent coverage, indicating rising global attention to the company.

India’s Startup Policy Forum Launches DeepVerse, Lists Quantum As A Priority

Startup Policy Forum’s DeepVerse will convene investors, founders and officials on deep-tech policy, with quantum technology among 10 focus areas.

A Deep Dive Into Grok 4.6’S Role In AI On Gemini Enterprise Agent Platform

xAI’s Grok 4.6, its flagship AI model, is now accessible through the Gemini Enterprise Agent Platform, expanding enterprise deployment options.

Court Denies Musk’s Attempt To Stop Minnesota’s AI Nudification Ban

A judge rejects Elon Musk’s bid to temporarily block Minnesota’s law banning AI tools that create nonconsensual nude images, allowing enforcement during legal proceedings.