Gemini 4 Neon: Google Launch and Business API Support
Google unveiled Gemini 4 Neon. Google AI Pro and Ultra subscribers can use it in the Gemini app, and businesses can access it through the API or model selection menu in Gemini for business.
Google AI Pro and Ultra subscribers can use Gemini 4 Neon in the Gemini app and Google Antigravity.
Businesses can use Neon through the API or model selection menu in Gemini for business.
Standard API pricing per 1 million tokens is $3 for regular input, $0.15 for cached input, and $15 for output.
In an independent evaluation by Artificial Analysis at launch, Neon scored 61 points with the xhigh setting, placing 3rd overall.
Google did not say when the rollout to Korean accounts would be complete or disclose the request limit specific to Neon in the Gemini app.
Google unveiled its new AI model, Gemini 4 Neon, on November 18, 2026. Google AI Pro and Ultra subscribers can use it in the Gemini app. Businesses can also access it through the Gemini for business API.
We checked Google's announcement and documentation, along with the Google DeepMind model card, as of November 20, 2026.
The default model for new conversations in Google Antigravity
According to Google's documentation, Neon is the default model for new conversations if you have never selected a model. If you have already selected one, Google Antigravity keeps that choice. So the model you see when you open a new conversation depends on your previous selection.
How Pro and Ultra subscribers and businesses can access Neon
Google DeepMind said Google AI Pro and Ultra subscribers can use Neon. Google also explained how to select Neon in its business services. It did not announce when the rollout to Korean accounts would be complete.
Who can use it | Service | How to access Neon
Google AI Pro and Ultra subscribers | Gemini app | Use it in the app
Paid Google AI Pro and Ultra subscribers | Google Antigravity | Use it in new conversations
Businesses | Gemini for business | API calls or the model selection menu
Businesses | Google Antigravity | Use it within the service
How much does the API charge for input and output tokens?
Google announced a standard API rate of $3 per 1 million regular input tokens. Cached input costs 95% less than regular input. Output generated by the model is charged at a separate rate.
API billing category | Rate announced on November 18, 2026
Regular input | $3 per 1 million tokens
Cached input | $0.15 per 1 million tokens
Output | $15 per 1 million tokens
3rd overall in Artificial Analysis's independent evaluation
At launch, Artificial Analysis ranked Neon 3rd overall on its Intelligence Index. It evaluated Neon with the xhigh setting. For comparison, it used the high setting for Gemini 4 Carbon.
Evaluation metric on November 18, 2026 | Neon result
Intelligence Index score | 61 points
Score difference versus Gemini 4 Carbon | Neon scored 1 point higher
Rank among all models evaluated | 3rd
No date for the end of the Korean rollout or details on app request limits
Google did not say when Neon would finish rolling out to all Korean accounts. It also did not disclose how many requests users can make to Neon in the Gemini app. The official announcement gives neither a publication time nor a time zone, so the time cannot be converted to Korean time.
What to check before using Neon with a Korean account
To use the Gemini app in Korea, check your account's subscription tier and model list. To calculate API costs, break down your usage by billing category.
· In the Gemini app, check whether your current account is covered by a Google AI Pro or Ultra subscription.
· Check whether Gemini 4 Neon appears in that account's model selection menu.
· If you use the business API, apply the respective rates to your regular input, cached input, and output usage.