ToolNavs Find Useful AI Tools
Submit Sign in
Back to AI information
Gemini 4 Neon Launches: Frontier Reasoning, Faster and Cheaper, Available Today for Pro and Ultra Users

Gemini 4 Neon Launches: Frontier Reasoning, Faster and Cheaper, Available Today for Pro and Ultra Users

AI information • Admin • • 26 views

On November 18, 2026, Google DeepMind announced Gemini 4 Neon on its official blog, packaging frontier-level reasoning into a faster, cheaper everyday flagship model. It sits between two existing models: Carbon handles the hardest long-horizon tasks, Hydrogen handles fast everyday execution, and Neon is meant to preserve both depth and efficiency in complex, high-frequency workflows.

Strongest in coding, with prices cut outright

On benchmarks, Neon scored 61 on the Artificial Analysis Intelligence Index at the xhigh thinking level, ranking third overall, behind only two larger, slower models; it ranks second on the Vals Index (70.9%) and second on the Agent Arena Index with 1,336 points. Coding is its home turf: it ranks first on the Artificial Analysis Coding Agent Index (58.6), scored 65.2% on Terminal-Bench 4.1, and 46.2% on FrontierCode 2.0 at the default setting. Pricing is the most direct selling point of this launch: $3 per million input tokens, $0.15 for cached input, and $15 for output, with a 1-million-token context window. Google says the default setting costs 32% less per task than Carbon, and 50% to 80% less than top competing frontier models.

Safety stepped up, and every channel live at once

Safety is Neon's other side. DeepMind says it recorded its lowest fabrication rate yet in internal testing, with a 0% cheating rate on Terminal-Bench 4.0 and a 0% attack success rate on Gray Swan's indirect prompt injection benchmark at the highest thinking level. A cyber-defense variant, Neon for cyber, launches alongside it, posting a best-in-class 76% on the hard tier of AutoBaxBench and opening first to trusted defenders in the Fairwind Program. Starting today, Neon is live in the Gemini API and Google AI Studio, as well as in Gemini for business and Antigravity; Google AI Pro and Ultra subscribers can select it from the model dropdown in the Gemini app. Carbon, its sibling in the Gemini 4 family, recently topped the math leaderboard, and the division of labor between the two is now taking shape: one takes on the hardest problems, the other carries the high-frequency daily workload.

Recommended Tools

More