Featured image: An aerial view of Google’s main campus in Mountain View, California (file photo). Photo: Austin McKinley, Wikimedia Commons, CC BY 3.0. Cropped to 16:9.
Google on Wednesday, Nov. 18, 2026, released Gemini 4 Neon, a model it says brings “frontier-level intelligence” to complex coding, knowledge work and science tasks at lower cost and higher speed than its flagship, Gemini 4 Carbon. Neon is priced at $3 per million input tokens and $15 per million output tokens, with a 1 million-token context window, Google said.
Where Neon fits in the Gemini 4 family
Google described Neon as the middle option in its lineup. “While Carbon tackles your hardest, long-horizon projects and Hydrogen handles fast everyday execution, Neon balances depth and efficiency,” the company wrote. Like Hydrogen, Neon offers four thinking levels: low, medium, high and extra high, 9to5Google reported.
In the Gemini app, Neon is available to Google AI Pro and Ultra subscribers and replaces both Gemini 4 Argon and the older Gemini 3.1 Pro, according to 9to5Google. Developers can use it through the Gemini API in Google AI Studio, and businesses can reach it through Gemini for business and Google Antigravity. The Decoder called it Google’s fourth new Gemini model in a month. It follows the rollout of Gemini 4 Hydrogen to all Google Search users earlier this week.
What Google says it can do
Google said Neon ranks third overall, with a score of 61, on the Artificial Analysis Intelligence Index at its highest thinking level, and second on the Vals Index at 70.9%. It said Neon ranks first on the Artificial Analysis Coding Agent Index at 58.6 and scores 65.2% on Terminal-Bench 4.1. Google also cited a 75.8% score on Agents’ Last Exam and a second-place debut, at 1336, on Arena’s new Agent Arena Index.
On cost, Google said that at its default high thinking level, Neon has a 32% lower cost per task than Carbon. A version tuned for security work, Gemini 4 Neon for cyber, will roll out to vetted defenders in Google’s Fairwind Program in the coming days. The company also said Neon refused just 1.98% of tasks on Terminal-Bench 4.0 while keeping Carbon-level safeguards against cyber and biological misuse.
By the numbers: how the price compares
At $3 and $15, Neon costs twice as much as Hydrogen, 9to5Google reported. Carbon and Argon are on promotional pricing of $2 and $10 until Nov. 30, after which they rise to $4 and $20. At that point Neon will be 25% cheaper than Carbon. Cached input costs $0.15 per million tokens, a 95% discount, Google said.

The new model is also far cheaper than OpenAI’s top model. OpenAI’s GPT-6.2 Astra, released a day earlier, costs $8 per million input tokens and $40 per million output tokens, according to OpenAI, so Astra’s list prices are about 2.7 times Neon’s, according to Boston Report calculations. Benchmarks and prices do not capture every difference, and the two companies measure performance in different ways.
Why it matters
Google has been pushing Gemini into health care, search and business software, including Gemini for Healthcare, with a Massachusetts insurer among early users. A cheaper model that Google says performs close to its flagship could make those products less expensive to run at scale. For more on the race among AI developers, see our artificial intelligence coverage.
Sources
- Google: Introducing Gemini 4 Neon
- 9to5Google: Google launches Gemini 4 Neon with Carbon quality at Flash speeds
- Google DeepMind: Gemini 4 Neon model card
- The Decoder: Google ships Gemini 4 Neon, its fourth new Gemini model in a month
Read next in Artificial Intelligence
Artificial IntelligenceGoogle Says It Found an AI Framework That Writes and Runs Exploits From Software Patches, With No Human in the Loop
Artificial IntelligenceOpenAI Releases GPT-6.2 Astra and Sol, Cutting Astra’s API Price 20%
Artificial IntelligenceDario Amodei: The Physicist Who Left OpenAI to Found Anthropic and Now Takes It Public
Most read this week
- Finland Confirms Strike on Russian Warship in Gulf of Finland; Ukraine’s SBU Claims Drone Attack
- U.S. Envoys Witkoff and Kushner Hold Talks With Putin in Moscow, Then Head to Dublin to Brief Zelensky
- Israeli Strike on Gaza’s Nasser Hospital Kills Six, Including a Surgeon, Health Ministry Says
- Nvidia Earnings, Walmart and Fed Minutes Headline a Busy Week for Wall Street: What to Watch Nov. 16-20
- Iran’s Ghalibaf Says Hormuz Will ‘Never Return’ to Pre-War Status as Another Tanker Is Hit
Latest
- Boston City Council Votes 8-4 to End Parking Minimums for New Housing
- Lee Accepts Takaichi’s Invitation to Visit Japan in January After Their Manila Summit
- Iran Fires Drones at a U.S.-Used Air Base in Kuwait After American Jets Strike Qeshm Island
- Kings Shut Out Bruins 3-0 as Boston Drops a Third Straight Game at TD Garden
- Nvidia Tops Forecasts With a $115.8 Billion Quarter, but Shares Slide in Its First Report Since Huang’s Exit
Get The Boston Report Briefing
Our free weekly email with the most important stories from Boston and the world. Unsubscribe anytime.
Free. We never sell your address. Privacy policy