Google brings Gemini 4 Neon to Pro subscribers
Google released Gemini 4 Neon on Nov. 18, its fourth Gemini 4 model in under two months. It is rolling out through Google’s API, AI Studio and enterprise products, with Pro and Ultra subscribers gaining access through the Gemini app and Antigravity. Its context window holds one million tokens.
Google’s model card shows Neon beating Claude Opus 5.5 and GPT-6.2 Sol in 13 of the 16 benchmark rows that include all three models. Reported strengths include long-context processing, automation and cybersecurity, although rivals lead on some tests. Google says Neon is adapted from Carbon and can still hallucinate, leaving developers to validate performance on their own workflows.
API pricing is $3 per million input tokens and $15 per million output tokens, 25% below Carbon’s announced standard rates. However, Carbon’s introductory rates of $2 for input and $10 for output remain available until Nov. 30, making Neon temporarily more expensive.
Independent measurements also highlight a reasoning-speed tradeoff. Artificial Analysis recorded an Intelligence Index score of 61 at xhigh reasoning and 60 at high, while estimated time per index task rises from 14.8 to 21.5 minutes. Those timings apply to benchmark tasks, not every workload. Google’s faster release cadence is visible, but how much AI contributed to the acceleration remains unclear.