# Gemini 4 Neon: release status, model ID and NEON tier

> Gemini 4 Neon is out: Google's November 18 launch, the gemini-4-neon model ID, 1M-token limits, API prices and the NEON tier question.

- Canonical HTML: https://omniakey.com/blog/gemini-4-neon
- Markdown: https://omniakey.com/blog/gemini-4-neon.md
- Published: 2026-11-16 (updated 2026-11-19)
- Author: Kiki, COO at OmniaKey
- Language: en
- Category: Guide
- Reading time: 7 min
- All posts: https://omniakey.com/blog/llms.txt

---

Gemini 4 Neon is available now. Google announced it on November 18, 2026 as a paid Gemini API model with a 1,048,576-token context window, a model ID labeled stable, and standard rates of $3 per million input tokens and $15 per million output tokens.

- **Available:** released November 18, 2026 on the paid Gemini API; no free tier.
- **Model ID:** `gemini-4-neon`, with 1,048,576 tokens of unified context and text output.
- **Price:** $3 per million input tokens and $15 per million output tokens at the standard tier.

**Updated November 19, 2026:** Google released Gemini 4 Neon on November 18; the status below now covers the shipped model.

## Is Gemini 4 Neon available today?

Yes. Google's [launch post](https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-neon/) is dated November 18, 2026, and the [AI Studio model page](https://aistudio.google.com/docs/models/gemini-4-neon) labels the model as stable. API calls run on the paid tier only: Google's [price page](https://ai.google.dev/gemini-api/docs/pricing) lists the free tier as unavailable for Neon.

Availability differs by product. The [Neon release notes](https://aistudio.google.com/docs/whats-new-gemini-4-neon) list both `generateContent` and the Interactions API. The [launch post](https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-neon/) names the Gemini app for Pro and Ultra subscribers, Antigravity for paid AI Pro or Ultra accounts, and Gemini for business in the API and model dropdown; Google Cloud's [Gemini 4 family page](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/start/get-started-with-gemini-4?hl=en) covers the Gemini Enterprise Agent Platform.

The November 16 sighting is now history: AI Tracker reported a gemini-4-neon string in the Gemini web app two days before Google's announcement.

## What does NEON mean in the Gemini 4 family?

Neon is one of four element names in the Gemini 4 line, alongside Hydrogen, Carbon and Argon. Google's [launch post](https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-neon/) positions Neon between Hydrogen, for fast everyday execution, and Carbon, for the hardest long-horizon projects; the post describes Neon as the option that balances depth and efficiency. Those are Google's product descriptions rather than independent measurements.

In Google's launch, the name describes the model rather than a subscription. The [Gemini 4 family](https://ai.google.dev/gemini-api/docs/models) now has four model IDs: `gemini-4-hydrogen`, `gemini-4-carbon`, `gemini-4-neon` and `gemini-4-argon`. For how the four arrived, see the [Gemini 4 release timeline](https://omniakey.com/blog/gemini-4-release-date).

## What are the model ID and limits?

The ID is `gemini-4-neon`. Google's [model page](https://aistudio.google.com/docs/models/gemini-4-neon) gives a unified context window of 1,048,576 tokens, which covers input, thinking and output together, and a default synchronous output cap of 65,536 tokens.

|Limit|Gemini 4 Neon|
|-|-|
|Context window|1,048,576 tokens|
|Default output|65,536 tokens|
|Long output|remaining context after input and thinking|
|Inputs|Text, image, video, audio, PDF|
|Output|Text|
|Knowledge cutoff|March 2026|

Long output uses the same context budget. Google's [release notes](https://aistudio.google.com/docs/whats-new-gemini-4-neon) say to set `background: true` and `max_output_tokens` above 65,536; without that setting the server stops at 65,536. Google sets the `max_output_tokens` ceiling at 1,048,576 minus input tokens, but that budget covers thinking and output together, so the visible answer is bounded by the context left after input and thinking.

Google's [release notes](https://aistudio.google.com/docs/whats-new-gemini-4-neon) list four thinking levels: `low`, `medium`, `high` (the default) and `xhigh`. `minimal` returns an error. Google's guidance is to start at `high`, use `xhigh` for hard multi-step work and test all four levels with a high output cap. Requests should drop `temperature`, `top_p` and `top_k`: the server ignores in-range values and returns a 400 for out-of-range ones.

Explicit caching is not supported; implicit caching runs on prompts of at least 4,096 tokens and discounts cached input by 95% on a hit, according to the [same release notes](https://aistudio.google.com/docs/whats-new-gemini-4-neon).

## What does Gemini 4 Neon cost?

Google's [price page](https://ai.google.dev/gemini-api/docs/pricing), checked November 19, 2026, lists flat standard rates for Neon: $3.00 per million input tokens, $15.00 per million output tokens and $0.15 per million cached input tokens. Output billing includes thinking tokens.

|Tier|Input|Output|Cached input|
|-|-|-|-|
|Standard|$3.00|$15.00|$0.15|
|Batch|$1.50|$7.50|$0.075|
|Flex|$1.50|$7.50|$0.075|
|Priority|$5.40|$27.00|$0.27|

Prices are in US dollars per million tokens. Pricing stays flat across the full 1M context window: Google's [release notes](https://aistudio.google.com/docs/whats-new-gemini-4-neon) say Neon has no long-context surcharge. Priority is 1.8 times the standard rate; Batch and Flex are half.

A request with 100,000 input tokens and 20,000 output tokens costs $0.60 at the standard rate. Input is 100,000 × $3.00 / 1,000,000 = $0.30, and output is 20,000 × $15.00 / 1,000,000 = $0.30. The same token counts cost $0.30 through Batch or Flex. These totals exclude grounding fees, which Google bills separately.

Caching changes the input side for repeated prompts. If the same 100,000-token prompt hits the implicit cache, input billing drops to $0.015, or 100,000 × $0.15 / 1,000,000, while the output charge stays the same.

## Where can you use Gemini 4 Neon?

|Place|Who can use Neon|
|-|-|
|Gemini API|Paid-tier projects|
|Gemini app|Pro and Ultra subscribers|
|Google Antigravity|Paid AI Pro and Ultra|
|Gemini for business|API and model dropdown|
|Gemini Enterprise Agent Platform|Global, US and EU endpoints|
|Gemini Managed Agents|Antigravity agent, alone or with Hydrogen|
|Fairwind Program|Trusted defenders, cyber variant|

The Gemini API is the main developer path: both generateContent and the Interactions API serve the model. In the Gemini app, Neon appears for Pro and Ultra subscribers; Antigravity limits it to paid AI Pro or Ultra accounts. The [launch post](https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-neon/) lists these channels.

The cyber variant is separate and gated. Google's [launch post](https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-neon/) says the Neon cyber variant is rolling out to trusted defenders through the Fairwind Program in the coming days. Google has not published a separate model ID for it on the pages checked.

## Is NEON a separate subscription tier?

The November 18 [launch post](https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-neon/) and the [API price page](https://ai.google.dev/gemini-api/docs/pricing) checked on November 19, 2026 do not announce a separate NEON subscription. The launch names Pro and Ultra subscriptions in the Gemini app, and the price page treats Neon as a model.

The November 16 sighting also included a NEON consumer tier string listed after Basic, Pro and Ultra; no official plan page, price, quota or benefit list for that tier appears in the Google sources checked.

The distinction matters for buyers: a Pro or Ultra subscription buys app access to Neon, while an API project pays per token. A NEON-branded plan would be a third arrangement; none appears in the launch or price pages checked. If a plan page appears, its price and included models will be the first things to check.

## How does Gemini 4 Neon compare with Carbon?

Google's [model card](https://deepmind.google/models/model-cards/gemini-4-neon/) says Neon is adapted from Carbon, so the two share architecture, training data, hardware and safety work. The practical differences are price, thinking defaults and the shape of the work each is built for.

Google describes Carbon as the model for the hardest long-horizon projects and Neon as the middle option that balances depth and efficiency. Neon defaults to `high` thinking and supports `xhigh`; Carbon defaults to `medium` and stops at `high`. A team moving a prompt from Carbon to Neon should set the thinking level explicitly rather than assume the same default.

The bigger difference is price. Neon's standard rate is $3.00 input and $15.00 output per million tokens. Carbon's [standard-tier rate](https://ai.google.dev/gemini-api/docs/pricing) is $2.00 input and $10.00 output per million tokens through November 30, 2026, then $4.00 and $20.00 from December 1. So Neon costs 50% more than Carbon during the promotional month and 25% less at Carbon's regular rate.

For the full side-by-side, see [Gemini 4 Neon vs Carbon](https://omniakey.com/blog/gemini-4-neon-vs-carbon).

## Where can you use a Gemini 4 model through OmniaKey?

OmniaKey lists its current Gemini models on [the Gemini models page](https://omniakey.com/models/gemini). For a Gemini 4 model with its own model page, start with [Gemini 4 Argon](https://omniakey.com/gemini-4-argon); the rate and limits on that page are the ones to check before switching. Current Gemini rates are also on [the pricing page](https://omniakey.com/pricing).

## Frequently asked questions

### Is Gemini 4 Neon free?

No. Google's price page lists Neon as paid-tier only, with no free-tier option. AI Studio access runs through a paid project.

### What is the Gemini 4 Neon model ID?

`gemini-4-neon`. Google's AI Studio model page labels it stable, and the same ID appears in the API model list.

### Can I use Gemini 4 Neon in the Gemini app?

Yes, if you have a Pro or Ultra subscription. Google's launch post names both tiers for the Gemini app and limits Antigravity access to paid AI Pro or Ultra accounts.

### Is there a NEON subscription plan?

Neither the launch post nor the API price page checked on November 19, 2026 announces a separate NEON plan. Those pages name Pro and Ultra subscriptions and treat Neon as a model; a separate NEON plan would need its own page, price and included-model list.

### How much does 1 million tokens cost?

At the standard tier, $3.00 for input and $15.00 for output. Batch and Flex cut both to half, and Priority charges 1.8 times the standard rate.

## Sources checked

- [Google: Introducing Gemini 4 Neon (November 18, 2026)](https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-4-neon/)
- [Google AI Studio: Gemini 4 Neon model page](https://aistudio.google.com/docs/models/gemini-4-neon)
- [Google AI Studio: What's new in Gemini 4 Neon](https://aistudio.google.com/docs/whats-new-gemini-4-neon)
- [Google AI Studio: Gemini API models](https://ai.google.dev/gemini-api/docs/models)
- [Google AI Studio: Gemini API pricing](https://ai.google.dev/gemini-api/docs/pricing)
- [Google Cloud: Get started with Gemini 4](https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/start/get-started-with-gemini-4?hl=en)
- [Google DeepMind: Gemini 4 Neon model card](https://deepmind.google/models/model-cards/gemini-4-neon/)

Checked November 19, 2026. Carbon's promotional rates run through November 30, 2026; recheck them on December 1.
