Google Gemini 4 Neon Launched: Price, Speed and Access
Google has launched Gemini 4 Neon, a new AI model offering near-flagship intelligence at a lower price and higher speed than its most powerful model. Announced on 18 November 2026 by Google DeepMind, Neon costs $3 per million input tokens and $15 per million output tokens through the Gemini API, and it is rolling out…
Google has launched Gemini 4 Neon, a new AI model offering near-flagship intelligence at a lower price and higher speed than its most powerful model. Announced on 18 November 2026 by Google DeepMind, Neon costs $3 per million input tokens and $15 per million output tokens through the Gemini API, and it is rolling out now to paid Google AI Pro and Ultra subscribers in the Gemini app, Google AI Studio, Gemini for business and Google Antigravity. It is not available on the free tier.
Neon is the fourth model in the Gemini 4 family, built for complex, high-frequency work such as software engineering, knowledge work and scientific analysis — jobs where users need strong reasoning but cannot wait for the slowest flagship model on every task. Its headline score is 61 on the Artificial Analysis Intelligence Index at its highest thinking level, third place overall.

Gemini 4 Neon at a glance
| Fact | Detail |
|---|---|
| Launch date | 18 November 2026 |
| API price | $3 per million input tokens; $15 per million output tokens |
| Cached input | $0.15 per million tokens (95% discount) |
| Context window | 1 million tokens |
| Intelligence Index | 61 at xhigh thinking level — third overall |
| Coding Agent Index | 58.6 — first place |
| Free tier | Not available — Pro or Ultra subscription required |
Checked on 19 November 2026.
Where Neon sits in the Gemini 4 family
Google has now released four Gemini 4 models in about five weeks: Argon (announced 30 September, generally available 13 October), Carbon (generally available 10 November), Hydrogen (12 November) and Neon (18 November). Google puts it plainly: Carbon tackles the hardest problems, Hydrogen handles fast everyday execution, and Neon balances depth and efficiency.
| Model | Launched | Role | Input / output per million tokens | Gemini app access |
|---|---|---|---|---|
| Argon | 13 Oct 2026 | Long-running coding | $2 / $10 through 30 Nov, then $4 / $20 | Pro and Ultra |
| Carbon | 10 Nov 2026 | Hardest projects | $2 / $10 through 30 Nov, then $4 / $20 | Ultra only |
| Hydrogen | 12 Nov 2026 | Everyday work | $1.50 / $7.50 | All tiers, including free |
| Neon | 18 Nov 2026 | Complex frequent work | $3 / $15 (flat) | Pro and Ultra |
In the Gemini app, 9to5Google reports that Neon takes over from Argon, and the older Gemini 3.1 Pro model is being retired. The picker now reads Auto, Everyday help (Hydrogen), Deeper analysis (Neon) and Code and complex work (Carbon). Price timing matters: through 30 November, Carbon and Argon cost less per token than Neon on introductory pricing, but on 1 December both rise to $4 and $20, making Neon the cheaper option per token as well as the faster one. Google says Neon at its default High thinking level already costs 32% less per completed task than Carbon.

Benchmarks, explained honestly
The numbers below come from Google’s announcement; third parties run the evaluations, but Google chose which results to publish. TimelyUps has not tested Neon.
The headline figure is the Artificial Analysis Intelligence Index: 61 at the highest (xhigh) thinking level, third overall and one point ahead of Carbon’s best published score of 60. On the Coding Agent Index, Neon ranks first with 58.6. Google also reports 65.2% on Terminal-Bench 4.1, 46.2% on FrontierCode 2.0, second on the Vals Index (70.9%), first on the Legal Research and Tax Agent benches and on Zapier’s AutomationBench (56.3%), and best in class on Terminal-Bench Science (70.5%), RiemannBench (84.0%) and HLE-Diamond (71.3%). It is second on Arena’s Agent Arena Index (1336).
Speed is the other half of the story. Google publishes no tokens-per-second figure, but independent coverage of Artificial Analysis data puts Neon at roughly 110 output tokens per second, about double Carbon’s roughly 57.
Gemini 4 Neon pricing
| Item | Price (USD) | Notes |
|---|---|---|
| Input tokens | $3 per million | Flat rate |
| Output tokens | $15 per million | $0.15 per million |
| Cached input | $0.15 per million | 95% off input rate |
| Context window | 1 million tokens | Text, images, audio, video |
| Carbon (until 30 Nov) | $2 / $10 | Promo — $4 / $20 from 1 Dec |
| Hydrogen | $1.50 / $7.50 | Half Neon’s price |
For developers, the comparison that matters is cost per completed task, not per token: Google says Neon costs 32% less per task than Carbon at High, and roughly 50% to 80% less than top competing frontier models.

How to access Gemini 4 Neon
Developers can build with Neon through the Gemini API paid tier in Google AI Studio and in Android Studio. Enterprises get it in Gemini for business and in Google Antigravity, where Neon can also act as an orchestrator directing parallel Hydrogen subagents. Consumers on a paid Google AI Pro or Ultra subscription can select Neon from the model dropdown in the Gemini app.
There is also a security-focused variant, Gemini 4 Neon for Cyber. Google says it produces more secure code than Carbon on AutoBaxBench, reaching state of the art on the hard tier at 76%, and it is rolling out to trusted defenders in the coming days through Google’s Fairwind Program. Read our report on the Java4Shell flaw that lets attackers run code.
What it means for Pakistan
Google announced Neon as a global rollout with no country restriction mentioned, and API pricing is in US dollars for developers everywhere. Pakistani developers using the paid Gemini API or AI Studio can therefore access Neon on the same terms — no Pakistan-specific price or launch date has been announced.
For Gemini app users in Pakistan, the key point is the subscription line: Neon is for Google AI Pro and Ultra subscribers, not free users. Anyone already paying for Pro gets a frontier-class model without moving to Ultra — and since Neon replaces Argon in the app’s Pro slot, Pro subscribers effectively get an upgrade at no extra cost. For startups, Neon is twice the price of Hydrogen but half the price of Carbon’s standard rate from 1 December, and it can be tested in AI Studio before production use. The AI race is moving fast: Anthropic has just priced its own IPO, covered in our report on the Anthropic IPO pricing at $83 a share, and launched new Claude models covered in our report on Claude Fable 5.2 and Mythos 5.2. One honest limit: Google has not announced any Pakistan-specific pricing in rupees, any local partnership, or any timeline for a free tier in Pakistan.

What is not yet known
Google has not said whether Neon will come to free Gemini users or to AI Mode in Search, where Carbon and Hydrogen are already available. No timing has been given for Neon on Vertex AI, and no parameter count or training details have been disclosed. The safety figures come from Google’s own evaluations and have not been independently verified.
The same lab recently used AI to verify Signal’s encryption with a 6.5-million-line maths proof, covered in our report on DeepMind’s AlphaProof Signal encryption verification. We will update this article when Google announces free-tier access, Vertex AI timing or independent benchmark results for Neon.
Frequently asked questions
What is Gemini 4 Neon?
Google’s newest AI model, launched on 18 November 2026. It sits between the fast, cheap Hydrogen and the flagship Carbon, offering near-flagship reasoning at higher speed and lower cost per task.
How much does Gemini 4 Neon cost?
$3 per million input tokens and $15 per million output tokens, with cached input at $0.15 per million. The price is flat; Carbon and Argon rise from $2/$10 to $4/$20 on 1 December 2026.
Is Gemini 4 Neon free?
No. In the Gemini app it is only for paid Google AI Pro and Ultra subscribers. Free-tier users stay on Hydrogen, and developers need the paid tier of the Gemini API.
How good is Gemini 4 Neon on benchmarks?
According to Google, Neon scores 61 on the Artificial Analysis Intelligence Index, third overall, and first on the Coding Agent Index with 58.6. Independent verification is still limited because the model is only a day old.
What is the difference between Neon, Carbon and Hydrogen?
Carbon is Google’s strongest model for the hardest projects; Hydrogen is the fast, low-cost model for everyday work; Neon sits in the middle, with near-Carbon capability at roughly double Carbon’s speed.
How do I use Gemini 4 Neon in Pakistan?
Pakistani developers can use it through the paid Gemini API in AI Studio on the same dollar pricing as everywhere else. In the Gemini app, paid Pro or Ultra subscribers can select Neon from the model dropdown.
