Google introduces Gemini 4 Neon for coding, knowledge work, and science
Original title: Gemini 4 Neon
Pricing is $3 per million input tokens and $15 per million output tokens, with a 1 million-token window.
Google’s announcement says Gemini 4 Neon is available today through the Gemini API, Google AI Studio, Gemini for business, paid Google Antigravity, and the Gemini app for Pro and Ultra subscribers. Google cites rankings on coding, agent, legal, finance, and science benchmarks and says Neon recorded its lowest fabrication rate yet in internal Antigravity testing. The post also describes a cyber variant used for secure-code generation, vulnerability finding, and patching.
Unconfirmed: Benchmark scores and cost savings versus competing models are stated only by Google.
- Google says Neon ranks #3 on Artificial Analysis with a score of 61 at xhigh.
- Cached input is listed at $0.15 per million tokens.
- Google reports #1 on the Artificial Analysis Coding Agent Index at 58.6.
- On WebExploitBench v1.2, Google claims a state-of-the-art 51.26%.
- Four thinking levels let teams trade capability against token cost.
- Vendors
- AI models
- Gemini 4 NeonHydrogen
From the original article
7 min read at blog.google
Bringing frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows. This content is generated by Google AI. Generative AI is experimental Today, we’re introducing Gemini 4 Neon — bringing frontier-level intelligence with fast end-to-end task completion across complex software engineering, knowledge work, and science workflows.
By yigitdemirag · blog.google
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.