Google Launches Gemini 4 Neon: Intelligence Index Score of 61 Ranks Third, Tops Coding Agent Index

  • Dispatched: Nov 19, 2026 · 13:38 ET (18:38 UTC)

One-sentence summary: Google DeepMind released Gemini 4 Neon on November 18, 2026 ET, a mid-tier model in the Gemini 4 family priced at $3 per million input tokens and $15 per million output tokens, scoring 61 on the independent Artificial Analysis Intelligence Index for third place overall and ranking first on the Coding Agent Index, with 32% lower cost per task than the flagship Carbon at its default setting.

Verification status: ✓ Official first-party disclosure (Google official blog)

Core Facts & Consensus Delta:

Google's official blog published Gemini 4 Neon on November 18, 2026 ET (the blog carries no exact timestamp; the earliest verifiable republication was 2026-11-18 16:26 ET / 21:26 UTC). Positioned between the flagship Carbon and the fast Hydrogen tier, Neon is the fourth Gemini 4 model released in about a month. It offers a 1-million-token context window, priced at $3 per million input tokens ($0.15 cached, a 95% discount) and $15 per million output tokens.

On benchmarks, Neon scores 61 on the Artificial Analysis Intelligence Index at its xhigh thinking level, ranking third overall behind larger, slower models; it ranks second on the Vals Index (70.9%) and first on the Coding Agent Index (58.6), with 65.2% on Terminal-Bench 4.1. On cost, Neon pairs its index score with a 32% lower cost per task than Carbon at the default high level, and sits on the Pareto frontier on both the Vals and Agent Arena indexes at roughly 50%-80% lower cost per task than top competing frontier models. A cyber variant, Neon for cyber, will roll out to trusted defenders in the coming days through the Fairwind Program. Neon began rolling out the same day in the Gemini API, Gemini for business, Google Antigravity and the Gemini app (Pro and Ultra subscribers). Google has not disclosed the model's training compute or parameter count.

Supply-Chain & Market Read-Through: Neon's direct competitive coordinates are OpenAI's GPT-6.2 Astra and Anthropic's Claude Fable 5.2, which sit ahead of it on the Intelligence Index; Neon enters high-frequency coding and agentic workloads with the top coding-agent score and lower cost per task, contesting the enterprise coding market on price and efficiency rather than flagship performance alone. Google (GOOGL) positions Neon as an orchestrator model inside Antigravity that directs parallel Hydrogen subagents, turning model tiering into compute scheduling inside its developer tool and concentrating inference demand on Google's own cloud; faster adoption would lift Gemini API and Google Cloud inference volumes and set a pricing reference for rival coding-model providers.

Next Catalysts & Watchpoints:

  • December 1, 2026: Carbon's promotional pricing ($2 input / $10 output per million tokens) expires; watch whether Google reprices the Gemini 4 tiers.
  • Independent benchmark updates and enterprise adoption data: watch Neon's ranking on third-party indexes and disclosures of large enterprise deployments.

Sources:

Information only, not investment advice.

→ 阅读中文版

More flash alerts from 2026-11-19

INSTANT ARCHIVE SEARCH

Search Stocks, Earnings & Macro Alerts

Quick tags: