🔍 Read the full analysis: Gemini 4 Neon: Reading The Fine Print On Its Gain And Lower Bill on ThorstenMeyerAI.com
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
TL;DR
Google released Gemini 4 Neon on November 18, positioning it as a balance of capability and efficiency. Artificial Analysis measured a one-point score gain over Gemini 4 Carbon at the highest settings, with lower cost and shorter task times; results vary by effort setting and benchmark. Google’s comparison table does not include the two models that lead the independent index.
Google released Gemini 4 Neon on November 18, presenting it as a model that balances capability and efficiency. In measurements from Artificial Analysis, Neon scored one point above Gemini 4 Carbon at the compared highest settings, while an index task cost less and took less time; the results do not show Neon leading the overall model rankings.
Artificial Analysis gave Neon at its xhigh effort setting an Intelligence Index score of 61, compared with 60 for Carbon at high. The organization estimated a task cost of $6.44 and 21.5 minutes for Neon xhigh, against $7.76 and 37.1 minutes for Carbon high at standard list prices. At the same high setting, Neon scored 60 and cost $3.74 per index task. These figures compare particular model settings, not every use case.
Neon’s listed price is $3 per million input tokens and $15 per million output tokens, below Carbon’s standard $4 and $20 rates. Carbon was discounted to $2 and $10 until November 30, according to the source report, so Neon is not cheaper than that temporary offer. Artificial Analysis attributed the measured cost difference to lower token prices and faster output, while noting Neon used about 15% more output tokens per task, reducing some of the savings.
The effort setting changes the trade-off. Artificial Analysis reported Neon scores and costs of 61 and $6.44 at xhigh, 60 and $3.74 at high, 55 and $1.90 at medium, and 44 and $0.75 at low. Google’s model card describes Neon as adapted from Carbon. It accepts text, images, audio and video, produces text, and has a one-million-token context window. Google lists a March 2026 knowledge cutoff and availability through several Gemini products and services.
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.
- 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.
- 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.
- 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.
- 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, Not a Clear Lead
For teams paying to run long or complex model tasks, the reported reduction in cost per index task and shorter completion time may matter more than a one-point score movement. At high effort, the index results suggest a less expensive way to reach Carbon’s high-setting score, while xhigh buys only one additional point at a noticeably higher task cost. Those are benchmark observations, not a guarantee of equivalent results on a company’s own prompts or workloads.
The comparisons also complicate claims of broad superiority. Artificial Analysis ranked Neon third among model families, behind GPT-6.2 Astra and Claude Fable 5.2. On its Coding Agent Index, Neon scored 58.6, close to Fable 5.2 at 58.5 and Claude Opus 5.5 at 58.2; the reported run cost was $6.86 for Neon, versus $15.47 and $12.61 for those Claude models. The close scores and different costs give developers a concrete reason to test models on their own coding tasks rather than rely on a single overall ranking.
There are counterexamples to the efficiency story. On Terminal-Bench 4.0, Neon xhigh scored 63.6%, below Carbon high at 65.2%. The index also reported a 14.4% hallucination rate for Neon on AA-Omniscience, compared with 11.4% for Carbon. Different benchmarks measure different capabilities, and these values should not be treated as a single verdict on reliability.
Products Worth Considering

DJI Mini 4K Drone with 4K UHD Camera, Under 249 g

Apple Watch SE 3 GPS + Cellular 40mm Smartwatch

Bose SoundTouch 30 Wireless Speaker, Black
As an affiliate, we earn on qualifying purchases.
A Rapid Gemini 4 Release Run
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, according to the source report. Google describes Carbon as suited to demanding, long-horizon projects and Hydrogen to fast everyday execution, with Neon between them. Its model card says Neon is adapted from Carbon and directs readers to Carbon’s documentation for architecture, training data and hardware details.
Google’s evaluation document reports Neon ahead in 16 of 19 comparison rows against Claude Opus 5.5 and GPT-6.2 Sol. Google itself lists competitors ahead in three: Claude Opus 5.5 on Terminal-bench 4.1 and FrontierCode 2.0, and GPT-6.2 Sol on Chartography. The comparison set does not include Claude Fable 5.2 or GPT-6.2 Astra, the two models that lead the cited Artificial Analysis index. That difference in test selection matters when interpreting Google’s table alongside independent rankings.
Google also said Neon achieved its lowest fabrication rate yet in internal Google Antigravity testing. That is a company-reported result using a different measurement from Artificial Analysis’s published AA-Omniscience figures; the source report says outsiders cannot independently check Google’s internal test. The two claims should not be treated as directly comparable.
“frontier-level intelligence with fast end-to-end task completion.”
— Google, in its launch post
Benchmark Scope and Pricing Caveats
The available figures do not establish how Neon will perform across different prompts, workloads or deployment conditions. The score comparison pairs Neon xhigh with Carbon high for one headline result, while other rows use different settings; effort level itself produces large changes in cost and score. The reported task costs are based on Artificial Analysis’s index and standard list prices, not a universal cost for every request.
Google’s 16-of-19 table covers a different comparison set from the independent ranking and uses its own evaluation document. The source material does not provide enough detail to reconcile every test or determine how results would change with the omitted leaders included under Google’s methodology. Google’s internal fabrication-rate claim also remains not independently verifiable from the information provided. Carbon’s temporary discount was reported as lasting through November 30, but longer-term pricing and any later changes are not established here.
Availability and Further Testing
Google lists Neon as available through the Gemini API in Google AI Studio, Gemini for business, Google Antigravity for paid AI Pro and Ultra subscribers, and the Gemini app for Pro and Ultra subscribers. Google said a cyber-defense version, Gemini 4 Neon for cyber, would reach eligible trusted defenders in its Fairwind Program over the following few days. That rollout is limited to eligible program members, not general availability.
The next useful evidence will come from tests that compare Neon and its rivals on the same tasks, settings and pricing basis, alongside user experience in production. Buyers can also track whether Google’s stated availability and Carbon’s time-limited discount change after November 30. Until more comparable results are available, the current data support a narrower conclusion: Neon offers a modest measured score gain over Carbon at the cited high-end comparison, with lower list pricing and faster index-task completion, but not a clear lead across models or benchmarks.
Key Questions
How much better did Gemini 4 Neon score than Carbon?
Artificial Analysis gave Neon xhigh an Intelligence Index score of 61, compared with 60 for Carbon high. The one-point difference reflects those specific settings, not a universal performance gap.
Is Neon cheaper than Gemini 4 Carbon?
Neon’s standard listed rates are $3 per million input tokens and $15 per million output tokens, below Carbon’s standard $4 and $20 rates. The source report says Carbon was temporarily discounted to $2 and $10 until November 30, 2026.
Does Neon lead the independent model rankings?
No. In the cited Artificial Analysis ranking, Neon is third among model families, behind GPT-6.2 Astra and Claude Fable 5.2.
Where can people access Gemini 4 Neon?
Google lists access through the Gemini API in Google AI Studio, Gemini for business, Google Antigravity for paid AI Pro and Ultra subscribers, and the Gemini app for Pro and Ultra subscribers.
What remains unverified?
Google’s claim that Neon had its lowest fabrication rate in internal Antigravity testing is not independently verifiable from the cited material. Results across benchmarks and real-world workloads also remain uncertain.
Source: ThorstenMeyerAI.com
