ModelGoogleConfirmed
Story 1 of 18 in the Wednesday 18 November editionGoogle launches Gemini 4 Neon at $3 and $15 per million tokens
Rolling out on 18 November 2026 across Google AI Studio, Google Antigravity, the Gemini app, GitHub Copilot and Cursor with a 1M context window and four thinking levels.

Google and Google DeepMind launched Gemini 4 Neon (gemini-4-neon) on 18 November 2026, slotting a frontier reasoning model between Gemini 4 Hydrogen and Gemini 4 Carbon for repository engineering, multimodal analysis and scientific tasks. Adapted from Carbon with a March 2026 knowledge cutoff, Neon processes text, image, video, audio and PDF inputs inside one unified 1,048,576-token window, returning up to 65,536 synchronous output tokens or 1,048,576 tokens through background long decode. Paid Gemini API billing is $3 per million input tokens, $15 per million output tokens including thinking tokens and $0.15 per million implicitly cached tokens, with no free tier.
Thinking levels, coding benchmarks and day-one IDE rollouts
Developers tune reasoning across low, medium, high (default) and xhigh levels on generateContent and the Interactions API. At default high effort, Neon cuts per-task token cost by 32% against Carbon; at xhigh, it ranks third overall on the Artificial Analysis Intelligence Index (61) and second on both the Vals Index (70.9%) and Arena’s Agent Arena Index (1336 across 4,200+ sessions).

Across agentic coding, Neon tops the Artificial Analysis Coding Agent Index (58.6) and reaches 39.0% on SWE-sweep inside Google Antigravity, more than triple its 12.8% mini-SWE score. Google’s benchmark table lists 65.4% on CursorBench 4.1, whereas Cursor reported 63.7% at default effort (up from 57.3% for Hydrogen). Neon rolls out across Google AI Studio, Gemini for business, Gemini Managed Agents, Google AI Pro and Ultra plans in Antigravity and the Gemini app, Cursor and generally available GitHub Copilot Pro+, Max, Business and Enterprise tiers.

Technical details
- Model code and token window
- gemini-4-neon; 1,048,576 unified input-output tokens (65,536 synchronous output / 1,048,576 long decode)
- Paid API pricing (per 1M tokens)
- $3 input, $15 output (including thinking tokens) and $0.15 implicit cache; no free API tier
- Thinking levels and consumption
- low, medium, high (default) and xhigh; supports Batch API, Flex inference and Priority inference
- Supported API capabilities
- Function calling, code execution, computer use, URL context, File search, Google Search and Maps grounding
Release details & context
Science, enterprise knowledge and defensive cybersecurity benchmarks
Beyond repository refactoring, Google reported first-place finishes for Gemini 4 Neon on Legal Research Bench, Tax Agent Bench, Zapier’s AutomationBench (56.3%), Andon Labs’ Vending-Bench 2 ($16,843), Terminal-Bench Science 0.1 (70.5%), RiemannBench (84.0%) and HLE-Diamond (71.3%), alongside 95.2% on 256k-to-1M GraphWalks, 90.5% on SEC-bench Pro and 51.3% on WebExploitBench v1.2.
Dual-use offensive cyber capabilities are gated to verified defenders through Google DeepMind’s Fairwind Program. In the Gemini API, Neon omits explicit context caching, native image generation, audio output and Live API streaming, while pairing with Hydrogen subagents inside Google Antigravity /swarm workflows.
Common questions
Does Gemini 4 Neon include a free tier or explicit context caching in the Gemini API?
No, gemini-4-neon is restricted to paid API accounts and Google AI Pro or Ultra plans, and supports implicit context caching at $0.15 per million tokens rather than explicit cache creation.
When does Google recommend choosing Gemini 4 Neon over Hydrogen or Carbon?
Google positions Hydrogen for high-speed everyday tasks, Neon for high-frequency repository engineering, multimodal analysis and science, and Carbon for the hardest long-horizon architecture and security work.
Sources
- PrimaryGoogle Blog: Introducing Gemini 4 Neonblog.google
- OfficialGoogle AI Studio Docs: What’s new in Gemini 4 Neonaistudio.google.com
- OfficialGoogle DeepMind: Gemini 4 Neon Model Carddeepmind.google
- OfficialGemini Developer API Pricingai.google.dev
- OfficialGitHub Changelog: Gemini 4 Neon in GitHub Copilotgithub.blog
- OfficialCursor Docs: Gemini 4 Neoncursor.com
- On XCursor (@cursor_ai) CursorBench 4.1 post on Xx.com
Google release timeline (36 launches tracked)
- Google details /teamwork proofs and opens 10k free researcher grants
Google Research details Lean 4 proofs, NMR chemistry loops and C++-to-Rust rewrites with Antigravity /teamwork and opens 10,000 researcher grants.
- Google rolls out interactive sites from sources in Gemini Notebook
Google rolls out interactive site generation in Gemini Notebook on the web, turning uploaded sources into stateful dashboards and study apps.
- Google makes Gemini Tasks free in six countries on Gemini 4 Hydrogen
Google expands Gemini Tasks free of charge across six countries on Gemini 4 Hydrogen with email RSVPs, phone quoting and Google Pay checkout.
- Google AI Studio adds custom domains and HD Android emulator builds
Google AI Studio adds direct custom-domain purchasing and hosting alongside HD audio Android emulation, Firestore sync and personal Play publishing.

