TL;DR
Google released Gemini 4 Neon on November 18, 2026. Artificial Analysis scores it one point above Gemini 4 Carbon at the highest tested settings, while its index task costs less and takes less time at standard prices. Google’s launch table does not compare Neon with the two model families leading that independent index.
Google released Gemini 4 Neon on November 18, presenting it as a model that balances capability and efficiency. In Artificial Analysis’s independent Intelligence Index, Neon scored 61 at xhigh, one point above Gemini 4 Carbon at high; the same source measured a lower cost and shorter completion time for its index task, though the compared models used different effort settings.
Get business pricing on tech for your team
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
Artificial Analysis reports that the index task cost $6.44 with Neon at xhigh, compared with $7.76 for Carbon at high, using standard prices. The measured task took 21.5 minutes for Neon and 37.1 minutes for Carbon. These results are specific to the index’s task and measurement setup; they do not establish that every use of Neon will cost less or finish faster.
At the same high effort setting, the source says Neon matches Carbon’s score for $3.74 per task. Neon’s listed 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 supplied source, so the standard-price comparison does not reflect that temporary offer.
Google’s launch table reports Neon ahead in 16 of 19 rows against Claude Opus 5.5 and GPT-6.2 Sol. The supplied independent measurements place Neon third among model families on the Intelligence Index, behind GPT-6.2 Astra and Claude Fable 5.2. The benchmark comparisons use different model lineups, so Google’s table does not show how Neon performs against those two independent index leaders.

Lower Costs Shape the Upgrade
The release’s clearest measured change is efficiency per index task, rather than a large score jump. A one-point increase over Carbon at the tested settings is small, while the reported cost and time differences could matter to organizations running lengthy model tasks. The size of any savings in real deployments will depend on the prompts, output volume, effort setting, and applicable prices.
Pricing also affects the decision. Neon’s standard token rates are lower than Carbon’s standard rates, but the current Carbon discount runs through November 30. Buyers comparing the models before that date may see a different cost relationship than the standard-rate figures suggest. The task-cost estimate and token prices measure different things and should not be treated as interchangeable.
Google’s benchmark table provides one view of Neon’s performance, but its selected competitors limit what readers can conclude about its standing against the strongest models in the independent index. For teams evaluating a switch, the release offers evidence of a potentially more efficient Carbon-derived option, while leaving head-to-head results against the index leaders unanswered.
Products Worth Considering

Jackery HomePower 3000 Portable Power Station, 3072Wh

Bose QuietComfort Wireless Noise Cancelling Headphones, Sandstone

Bose QuietComfort True Wireless Earbuds with Active Noise Cancellation, Bluetooth 5.3, USB-C Charging, IPX4, Black
As an affiliate, we earn on qualifying purchases.
Google’s Rapid Gemini 4 Rollout
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-running projects and Hydrogen as aimed at fast everyday execution, with Neon positioned between them.
Google’s model card says Neon is adapted from Carbon and refers readers to Carbon documentation for architecture, training data, and hardware details. Google lists support for text, images, audio, and video inputs, text output, and a one-million-token context window. The model card gives a March 2026 knowledge cutoff and four effort levels, from low to xhigh, with high as the default.
According to the supplied launch information, Neon is 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. A cyber-defense version is being made available to trusted defenders in Google’s Fairwind Program, with access for eligible members expected over the following days.
““frontier-level intelligence with fast end-to-end task completion.””
— Google, in its November 18 launch post
Limits of the Benchmark Comparison
The supplied results do not provide a direct comparison between Neon and GPT-6.2 Astra or Claude Fable 5.2, the two families ranked ahead of it in the independent index. Google’s table instead compares Neon with GPT-6.2 Sol and Claude Opus 5.5. The available figures therefore do not settle how Neon fares against the index leaders under matched conditions.
The cost and time figures describe one index task at standard prices. The source does not establish that these results generalize to every workload, and the headline comparison uses different effort settings: Neon at xhigh and Carbon at high. The supplied material also does not give a complete account of Neon’s architecture, training data, or hardware; Google’s model card points to Carbon’s documentation for those details.
Pricing and Comparisons to Watch
Google’s stated availability includes its API, business offering, and eligible paid consumer subscriptions. Members eligible for the Fairwind Program were expected to receive access to the cyber-defense version over the days following the launch. The supplied material does not give a later access update.
The immediate pricing date to watch is November 30, when Carbon’s stated discount ends. Further independent measurements, especially matched tests against GPT-6.2 Astra and Claude Fable 5.2, would help clarify Neon’s relative performance and whether its measured task-cost advantage holds across other workloads.
Key Questions
When did Google release Gemini 4 Neon?
Google released Gemini 4 Neon on November 18, 2026, according to the supplied launch information.
How did Neon score against Gemini 4 Carbon?
Artificial Analysis scored Neon at 61 at xhigh and Carbon at 60 at high. Because the effort settings differ, this is not a same-setting comparison.
Was Neon cheaper and faster in the independent test?
For the Artificial Analysis index task, Neon at xhigh was measured at $6.44 and 21.5 minutes, while Carbon at high was measured at $7.76 and 37.1 minutes, using standard prices. The figures describe that task and setup, not every workload.
Does Google’s table compare Neon with the independent index leaders?
No. Google’s table compares Neon with Claude Opus 5.5 and GPT-6.2 Sol. The supplied independent index ranks GPT-6.2 Astra and Claude Fable 5.2 ahead of Neon, but does not provide the same Google-table comparison against them.
When does Carbon’s stated discount end?
The supplied source says Carbon’s discounted rates of $2 per million input tokens and $10 per million output tokens apply through November 30, 2026. It gives Neon’s standard rates as $3 and $15, respectively.
Source: Thorsten Meyer AI
