In this article
Google DeepMind releases Gemini 4 Neon, its fourth model in seven weeks
Google DeepMind has released Gemini 4 Neon, a model that runs faster and costs less than its predecessor, Gemini 4 Carbon, despite being released only weeks later. Neon achieves roughly 110 tokens per second, nearly double the rate of Carbon. It also ranks third on the Artificial Analysis Intelligence Index, trailing only OpenAI’s GPT-6.2 Astra and Anthropic’s Claude Fable 5.2.
The new model offers a one-million-token context window covering text, images, audio and video. Google positions Neon between Carbon and Hydrogen in its product hierarchy.
Pricing is lower because of list price, not efficiency
Cost savings come from the base list price rather than improved token efficiency. Through Google AI Studio and Gemini for Business, one million input tokens costs three dollars, or $0.15 when cached. One million output tokens cost fifteen dollars. This is 25 percent below Carbon’s standard price of four dollars and twenty cents, though it remains 50 percent higher than Carbon’s promotional rate, which continues through the end of November.
According to Artificial Analysis, per-task costs for Neon at the xhigh reasoning level reach $6.44. This is roughly 17 percent lower than Carbon’s undiscounted figure of $7.76. Neon actually uses about 15 percent more output tokens per task than Carbon, so the reduction in expense relies entirely on the lower per-token list price. At the high level, Neon matches Carbon’s intelligence score while cutting per-task costs in half. Processing time for the benchmark also drops from about 36 minutes to roughly 21 minutes because Neon generates around 110 tokens per second, nearly double Carbon’s rate of 57 tokens per second.
Coding agents run cheaper and visual analysis improves
The model performs competitively on the Coding Agent Index. Inside Google Antigravity, Neon at xhigh scores 59 points. This beats Carbon (55 points) as well as Claude Fable 5.2 and Claude Opus 5.5 (58 points each) at a fraction of the price. An average coding task costs $6.86 with Neon, compared to $12.61 for Opus and $15.47 for Fable. In the SWE-sweep test, the Antigravity environment triples Neon’s hit rate to 39.0 percent, according to Google.
Artificial Analysis measured large gains in visual processing. On the document test GDP.pdf, Neon at high hits the top score of 33.4 percent, five points ahead of Carbon, though it falls back to 28.0 percent at xhigh. On the AA-Briefcase test, Neon takes the lead for best presentation quality with 1771 Elo points, ahead of Opus 5.5.
In Google’s own evaluations, Neon leads individual categories on the Vals Index, Zapier’s AutomationBench (56.3 percent) and science tests such as RiemannBench (84.0 percent). For security professionals, Google is launching a cyber variant through the Fairwind Program that delivers top scores on SEC-bench Pro (90.5 percent) and WebExploitBench v1.2 (51.26 percent) at half the price of competing models. According to Neon’s Model Card, the model blocks 100 percent of indirect prompt-injection attacks on the Gray-Swan benchmark.
What it means
Developers using Google’s infrastructure see immediate benefits. Coding tasks run faster and cost less, while visual analysis capabilities improve on specific benchmarks. The shift to a lower base price means projects requiring high token counts become more affordable, even if the model consumes slightly more tokens per task than its predecessor.




