Google overtook Claude - Gemini 4 Neon was introduced, which showed better results than Opus 5.5 and GPT-6.2 Sol in most tests and tasks. What the new product can do: It is strongest in programming, logical reasoning and agent-based scenarios. In some measures it even surpasses the older Gemini 4 Carbon and takes the first place in the rating of coding agents. On an average level of thinking, Neon costs 32% less than Carbon. The model is already available through the API and Google AI Studio. The cost is $3 per million input tokens and $15 per million output tokens. After Claude's bans, we sit here.
1,337.