Gemini 4 Neon
3 storiesGemini 4 Neon is a frontier reasoning model from Google DeepMind with multimodal input and a 1M-token context, aimed at coding, research and agent workflows. Recent coverage notes it scores 61 on the Artificial Analysis Intelligence Index at its xhigh thinking level, ranking third, and places second on both the Vals Index and the Agent Arena steerability index. It is priced at $3 per million input tokens and $15 per million output tokens.
Related topics
OpenAI released GPT-6.2 Astra and Sol, both reaching its "Critical" cybersecurity threshold that previously limited such capability to vetted security teams, and added a Control Room monitoring system. Google's Gemini 4 Neon tops legal and tax benchmarks at $3 per million input tokens, less than half the $8 charged for OpenAI and Anthropic flagship models.
Frontier reasoning model for faster coding, research and agent workflows with multimodal and 1M-token context.
Gemini 4 Neon scores 61 on the Artificial Analysis Intelligence Index at its xhigh thinking level, ranking third behind larger, slower models; it ranks second on the Vals Index and second on the Agent Arena steerability index. Priced at $3 per million input tokens and $15 per million output tokens, with a 95% cache discount and a 1M-token context window, it rolls out today via the Gemini API and Google AI Studio.
That is everything