What changed
Google DeepMind announced the release of Gemini 4 Neon, a model that ranks #2 on the Vals Index and #3 overall on the Artificial Analysis Intelligence Index. It is designed for complex software engineering, knowledge work, and science workflows, with a focus on fast end-to-end task completion. Neon is priced at $3 per million input tokens and $15 per million output tokens, with a 1 million-token context window.
What this means in practice
Gemini 4 Neon is available through the Gemini API in Google AI Studio, Gemini for business, and for paid Google AI Pro and Ultra users. It excels in multi-step instruction following and has a low fabrication rate in internal testing. The model is also noted for its performance on benchmarks like the Legal Research Bench and Tax Agent Bench, where it ranks #1. Neon is capable of handling complex tasks such as building a photorealistic 3D lab bench and creating a first-person WebGL2 ski simulator from scratch.
Limitations and availability
Gemini 4 Neon is rolling out today in the Gemini API and is available for enterprise, analytical, and scientific work. It is priced at $3 per million input tokens and $15 per million output tokens. The model is available for use in Google AI Studio, Gemini for business, and for paid Google AI Pro and Ultra users. The supplied announcement does not specify the exact hardware requirements or licensing conditions for using the model.
![Source facts for Google DeepMind Launches Gemini 4 Neon with Enhanced Performance: Koray Kavukcuoglu SVP, Google DeepMind and Chief AI Architect, Google Share Your browser does not support the audio element. Listen to article [[duration]] minutes This content is generated by Google AI. On the Vals Index, Neon ranks #2 and lands on the Pareto frontier.](/images/articles/google-deepmind-launches-gemini-4-neon-with-enhanced-performance-editorial-facts.png)