Skip to main content
Live
Main content

Google launches Gemini 4 Neon at 32% lower cost per task than Carbon

The new mid-tier model costs $3 per million input tokens, carries a 1M-token context window, and ranks #2 on the Vals Index.

Jaeden Schafer
Editor in Chief · · 5 min read
Google logo

Google launched Gemini 4 Neon on November 18, a new model in the Gemini 4 family that the company says delivers frontier-level performance at 32% lower cost per task than Gemini 4 Carbon at its default high thinking level. Neon is priced at $3 per million input tokens and $15 per million output tokens, with a 1 million-token context window, and began rolling out the same day across the Gemini API, Google Antigravity and the Gemini app.

Google is positioning Neon between its two other new Gemini 4 models. In a blog post by Koray Kavukcuoglu, SVP at Google DeepMind and Google's Chief AI Architect, the company described Carbon as the model for the hardest long-horizon projects and Hydrogen as the one for fast everyday execution, with Neon balancing depth and efficiency. Carbon and Hydrogen both shipped the previous week.

On third-party leaderboards, Google says Gemini 4 Neon ranks #3 overall on the Artificial Analysis Intelligence Index with a score of 61 at its xhigh thinking level, just behind larger, slower models. It ranks #2 on the Vals Index at 70.9% and #2 globally on Arena's Agent Arena Index with a score of 1336 across more than 4,200 real-world agent sessions.

Key facts

  • 01Google released Gemini 4 Neon on November 18, 2026, rolling out in the Gemini API, Google Antigravity and the Gemini app.
  • 02Neon is priced at $3 per million input tokens and $15 per million output tokens, with a 1 million-token context window.
  • 03Google says Neon costs 32% less per task than Gemini 4 Carbon at the default high thinking level.
  • 04Neon ranks #3 on the Artificial Analysis Intelligence Index (61), #2 on the Vals Index (70.9%) and #1 on the AA Coding Agent Index (58.6).
  • 05A cyber version is rolling out to vetted defenders through Google's Fairwind Program in the coming days.

Coding is where Neon posts its strongest numbers. Google reports a #1 overall score of 58.6 on the Artificial Analysis Coding Agent Index, 65.2% on Terminal-Bench 4.1 and 46.2% on FrontierCode 2.0 at its default high setting. On SWE-sweep, Neon leads the mini-SWE harness at 12.8% and reaches 39.0% inside Google Antigravity, where it can also act as an orchestrator directing parallel Hydrogen subagents.

Google also lists enterprise and science results. Neon ranks #1 on Legal Research Bench, Tax Agent Bench and Zapier AutomationBench (56.3%), and tops Andon Labs Vending-Bench 2 with $16,843. In science and math, it scores 70.5% on Terminal-Bench Science, 84.0% on RiemannBench and 71.3% on HLE-Diamond, all of which Google calls best in class.

Cost is the core pitch. At the high thinking level, Google says Neon runs at roughly 50% to 80% lower cost per task than top competing frontier models, which it did not name. At xhigh, Google says Neon still costs at least 36% less per task than competing models at their highest settings. Cached input tokens are priced at $0.15 per million, a 95% discount.

Neon's list price is twice that of Hydrogen. Gemini 4 Argon and Carbon are priced at $2 per million input tokens and $10 per million output tokens until November 30, after which they rise to $4 and $20, making Neon 25% cheaper than those models once the promotional pricing ends.

In the Gemini app, Neon is available to Google AI Pro and AI Ultra subscribers and replaces Argon in the model picker. The updated picker lists an Auto option, Hydrogen for everyday help, Neon for deeper analysis and Carbon for code and complex work, while Gemini 3.1 Pro is being retired.

Related · from this week
Nvidia launches $4,999 DGX Spark 64GB for local AI development
Jaeden Schafer · 4 min read →

Google is also releasing a security-focused version. Gemini 4 Neon for cyber is rolling out to trusted defenders through the Fairwind Program over the next few days, with access requests handled by Google Cloud account teams. Google reports a state-of-the-art 51.26% on WebExploitBench v1.2, 76% on the hard tier of AutoBaxBench and 70.7% on Collinear's CWE-bench v2, a 3.4% uplift over Carbon.

On safety, Google reports a 0.09% misaligned action rate on its agentic safety suite, a 99.2% pass rate on its agent honesty evaluation and a 0.0% attack success rate on the Gray Swan indirect prompt injection benchmark at Neon's highest thinking level. The company says Neon also posted its lowest fabrication rate yet in internal Google Antigravity testing.

Developers can access Gemini 4 Neon on the paid tier in Google AI Studio and Android Studio, and all Pro and Ultra subscribers can try it in AI Studio. Business customers can reach it through the API and the Gemini for business model dropdown, and Google has published a model card alongside the launch.

ShareXLinkedInEmail
AI Box

Every AI model. One chat.

The latest models from ChatGPT, Claude, Gemini, Sora, ElevenLabs — 80+ models in a single chat. Compare answers side by side. Pick the best one every time.

  • ChatGPT, Claude, Gemini, Grok, DeepSeek — in one chat
  • Generate images & video with Sora, Veo, Ideogram
  • Compare any two models side by side
  • From $8.99/mo · 80+ models, all included
Try AI Boxaibox.ai
Trusted by 3,000+ teams
Got a tip?

Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.

Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.

AI Box Daily briefingFree · Daily · No fluff

Stay ahead of everyone in AI.

The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.

Loved by 10,000+ AI professionals
Free forever. Unsubscribe with one click.

The briefing read inside teams at

Keep reading

More from Models

Nvidia logo
Models

Nvidia launches $4,999 DGX Spark 64GB for local AI development

The new SKU runs 100B-parameter models on device; two units cluster to 128GB and 200B via Nvidia Sync Cluster Assistant.

Jaeden Schafer4 min read
OpenAI logo
Models

OpenAI's Astra and Anthropic's Claude Opus 5 crack unsolved Enigma messages

Two cryptanalysts used frontier LLMs to break wartime codes that had resisted human researchers for two decades.

Jaeden Schafer4 min read
Anthropic logo
Models

Anthropic says Claude autonomously discovered a Crispr-like enzyme system

Nearly 950 Claude agents worked 21 hours across 210 million tokens to flag a previously uncharacterized enzyme found in bacteriophages.

Jaeden Schafer4 min read