AI · 19 November 2026
Google Gemini 4 Neon: fourth new Gemini model in a month
Google has released Gemini 4 Neon, a frontier model accepting text, images, video, audio and PDFs, with a one-million-token context window and listed pricing of $3 and $15 per million tokens.
What happened
Google has released Gemini 4 Neon, a new frontier model designed for software development, knowledge work and science. It is the fourth new Gemini model the company has shipped within a month, according to The Decoder.
Neon accepts text, images, video, audio and PDF documents. It supports a context window of one million tokens, so it can work across large volumes of material in a single session.
Pricing sets Neon apart from earlier Gemini 4 models. It is listed at $3 per million input tokens and $15 per million output tokens, which The Decoder reports as higher than the previous Gemini 4 models.
Why it matters
The significance lies less in any single specification than in what a model of this breadth makes practical. A system that reads documents, interprets video and audio, and holds a million tokens of context can take on tasks that once required several specialised tools and manual hand-offs. For organisations, that points towards workflows where one model supports a support agent, an analyst and an engineer, rather than a patchwork of point solutions.
The pace of release matters as well. Four Gemini models in a month shows how quickly the baseline of available capability is moving, which puts pressure on operations teams to keep evaluating, testing and retiring tools. Pricing is the other lever. A higher-priced model prompts leaders to decide where extra capability justifies its cost and where a less expensive option is sufficient.
By the numbers
- 1 million tokens: the context window supported by Neon.
- $3 per million: the listed price for input tokens.
- $15 per million: the listed price for output tokens.
- Fourth: the number of new Gemini models released within a month.
The Renascence take
Most coverage will centre on specifications and token pricing. The more useful question for CX and operations leaders is how little of the value sits in the model itself once capability becomes widely available.
Teams will feel pressure to chase each new release, but a model is only one component of the customer experience. The behavioural trap is novelty bias: adopting the latest tier because it is new, rather than because a specific journey is failing. Start with the moments where customers wait, repeat themselves or get passed between teams, then test whether Neon's extra capacity removes that friction at a cost that is justified. Treat price as a design constraint, not a footnote.
Sources
This briefing was written by our Newsdesk, synthesising reporting from the outlets below. Follow the links for the original coverage.
FAQ
Questions we get on this topic
More in AI
Stay ahead of CX
Get the signal, not the noise.
The stories shaping customer experience — plus the Journal and Experience Loom — in your inbox.


