Product update

Google launches Gemini 4 Neon long-context frontier model

Event time 3 independent sourcesEditorial score 86/100Updated here

The short version

Google launched Gemini 4 Neon, a distinct Gemini 4 model for complex coding, analysis, and enterprise workflows, with a 1M-token context window, separate API pricing, and paid API, business, and subscriber rollout paths.

What changed

Google added Gemini 4 Neon at $3, $0.15 cached, and $15 per million input, cached-input, and output tokens, and announced phased availability through the Gemini API, Gemini for business, Google Antigravity, and eligible Gemini app users.

What it means for you

This is a model launch with its own availability and price, not a feature update to an existing tool; teams need to reassess quality, latency, cost, and access boundaries for long-context multimodal work.

Google launched Gemini 4 Neon. It is a separately named and priced model in the Gemini 4 line for complex software engineering, knowledge work, and science workflows, with a 1 million-token context window. It is not a feature update to an existing tool, so it changes the model-selection options for teams weighing quality, cost, and access.

Google lists standard API pricing of $3 per million input tokens, $0.15 cached input tokens, and $15 output tokens. The company says Neon is rolling out to the paid Gemini API tier in AI Studio, Gemini for business, Google Antigravity, and eligible Google AI Pro and Ultra subscribers in the Gemini app. Check current console availability, regional coverage, and plan limits before committing a deployment.

Long context is not automatically trustworthy context. Evaluate the model first on redacted, reversible work, such as producing a reviewable brief from PDFs and spreadsheets or proposing a tested patch in an isolated repository. Record total task tokens, latency, retries, and human review cost. Keep least-privilege tooling, source verification, and approval gates around workflows that call tools or read internal data. Google's benchmark and steerability results are useful screening signals, but do not replace validation on representative work.

Related directory entry

Fact check

  • VerifiedGoogle released Gemini 4 Neon on November 18, 2026 and positions it for complex software engineering, knowledge work, and scientific workflows.Evidence
  • VerifiedGoogle lists Gemini 4 Neon at $3 per million input tokens, $0.15 cached input tokens, and $15 output tokens, with a 1 million-token context window.Evidence
  • VerifiedGoogle says Neon is rolling out to the paid Gemini API tier in AI Studio, Gemini for business, Google Antigravity, and eligible Google AI Pro and Ultra Gemini app users.Evidence

Coverage timeline

  1. Primary sourceGoogle
    Google
  2. Primary sourceGoogle DeepMind
    Google DeepMind
  3. Primary sourceGitHub
    GitHub