Google releases new AI model Gemini 4 Neon
The model will become available in stages from 18 November in Google's apps, development environment and business services.
Listen to this article
There is no audio version yet. Request one and an AI voice will read the article aloud.
Read by an AI voice.
Google has announced Gemini 4 Neon and made it immediately available to several paid services and developers. The company positions the model as a new variant within the Gemini 4 family, but independent comparisons with competitors remain limited.
According to Google, Gemini 4 Neon will roll out from 18 November in the Gemini API, Google AI Studio, Gemini for business and the Gemini app for paying Pro and Ultra subscribers. Availability may vary by service and region. Google also published a model card for the system.
The model card describes, among other things, the capabilities tested and the safety outcomes. For text-to-text, multilingual tasks and image-to-text, Google says the differences from Gemini 4 Hydrogen are small enough to fall within the margins of uncertainty. This describes Google's own evaluations, not an independent ranking.
Google links Neon to applications for coding, analytical reasoning and tasks requiring multiple steps. As an example, the company gives an interactive reconstruction of a historical study using archival images. Such examples show what the software can generate, but in themselves say little about reliability in everyday use.
The launch follows shortly after other new models from Google and fits the acceleration of the AI market. For developers, what matters most is which access, pricing, processing limits and terms apply. Companies also face questions about data use, copyright, security and the verifiability of results.
The first independent reporting confirms the launch, but no broad comparative assessment is yet available that establishes its performance against other models. Users should therefore distinguish marketing claims from the results in the model card and from experiences in specific applications. Availability has been confirmed; whether the model represents a breakthrough in AI capability requires independent comparison.
Also show the region and a close-up
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The text is based on Google's announcement, the official model card and an independent technical news source. Marketing claims are presented as claims by Google, not as independent findings.
- confirmed Gemini 4 Neon was rolled out from 18 November 2026. — This is stated in Google's official announcement. source
- confirmed The model is available in, among other places, the Gemini API, Google AI Studio and paid Gemini services. — Google explicitly names these distribution channels. source
- confirmed The model card reports performance comparable with Gemini 4 Hydrogen on certain tasks within margins of uncertainty. — This is stated in Google's official DeepMind model card. source
- uncertain No broad independent comparison is yet available. — The independent reporting consulted confirms the launch but contains no broad comparative evaluation; this may change quickly. source
Editor's note
The launch and availability have been confirmed by Google. Performance claims are limited to what Google itself reports in the model card; independent comparisons remain scarce.Sources
- Introducing Gemini 4 Neon — Google
- Gemini 4 Neon — Model Card — Google DeepMind
- Google、「Gemini 4 Neon」を発表 — Impress Watch
More on this in Dutch media
- RTL Nieuws — „google gemini”
- FD — „google gemini”
- Trouw — „google gemini”