Google Launches Gemini 4 Neon, a Fast Frontier Model for Agentic Coding and Enterprise Work
Google released Gemini 4 Neon, a mid-tier frontier model with a 1M-token context window that ranks at or near the top of many agentic, coding, scientific and enterprise benchmarks. It does so at substantially lower cost per task than competing models. The model can orchestrate parallel subagent swarms for long-running tasks, and a cyber variant with state-of-the-art penetration testing and code hardening performance is rolling out to trusted defenders.
Skynet Chance (+0.04 pp): Cheaper, highly capable long-horizon agents that orchestrate subagent swarms and achieve state-of-the-art offensive exploitation scores broaden autonomous AI action and misuse potential. Stated alignment and safety improvements partially offset this.
Skynet Date (1 day sooner): Lowering the cost of frontier agentic capability speeds widespread deployment of autonomous systems.
AGI Progress (+0.03 pp): The model shows strong gains on long-horizon agentic tasks, multi-step reasoning, scientific problems and reduced fabrication, with frontier performance at much lower cost. These advance general-purpose autonomous work.
AGI Date (1 day sooner): Large efficiency gains and rapid tiered model releases indicate accelerating iteration, bringing capable agents to scale sooner.
<< All AI news for November 18, 2026
[ Get the daily index digest on Telegram → ]Related AI News
- Nadella Warns Unpatchable Critical Infrastructure Is Exposed to AI-Aided Cyberattacks
- Meta's Muse AI Agent Accessed Private Messages and Behaved Erratically, Internal Chats Reveal
- Microsoft Positions Copilot as 'New OS for Work,' Launches Autonomous Autopilot Agents With Governance Controls
- Anthropic IPO Filing Reveals Amazon and Google Will Hold Nearly 30% Combined Stake
- Google Rolls Out Gemini Tasks Agent Mode Free to Hundreds of Millions of Users