Google DeepMind rolled out three new Gemini models on Tuesday, expanding its AI portfolio with a focus on efficiency, speed and niche security functions. The three additions—Gemini 3.6 Flash, Gemini 3.5 Flash‑Lite and Gemini 3.5 Flash Cyber—are positioned as lower‑cost alternatives to the company’s flagship Gemini Pro line.
Gemini 3.6 Flash, billed as the "workhorse" of the series, delivers better coding assistance, knowledge‑work handling and multimodal performance while cutting token usage by as much as 17 percent. The reduction translates into cheaper operation compared with the earlier 3.5 Flash model, a claim Google says will benefit developers and enterprises that run AI agents at scale.
At the opposite end of the price spectrum, Gemini 3.5 Flash‑Lite offers the most economical option in the new lineup. Google markets the Lite variant as the go‑to choice for high‑volume workloads that demand rapid responses without the premium price tag.
Gemini 3.5 Flash Cyber targets a narrower audience. Fine‑tuned to locate and remediate cybersecurity vulnerabilities, the model is being released through a limited‑access pilot for governments and trusted partners. Google positions the Cyber model as a cost‑effective tool for defensive security operations, though broader commercial availability is not announced.
The three releases underscore Google’s strategy to deliver "efficiency, latency and reliability" for customers building AI agents in production. By offering models that trade raw capability for speed and price, the company aims to capture workloads that cannot justify the higher expense of Gemini Pro.
Gemini Pro, Google’s highest‑capability offering for complex reasoning and advanced coding tasks, did not receive an update in this cycle. The last refresh to the Pro line occurred in February, and the anticipated 3.5 Pro version remains delayed. Bloomberg reported internal setbacks as the team works to meet performance targets, and Google has hinted that the Pro model is still being tested with partners and may launch later this year.
Competitors have accelerated their own release schedules. OpenAI introduced GPT‑5.5 and began rolling out GPT‑5.6, while Anthropic launched Claude Opus 4.8, Claude Sonnet 5 and expanded access to its frontier Fable 5 model. Google’s announcement therefore arrives amid a crowded field of generative‑AI upgrades.
Logan Kilpatrick, product lead for Gemini at DeepMind, said the team is currently testing Gemini 3.5 Pro with select partners and hopes to "land soon." He also noted that Google has begun its most ambitious pre‑training run yet for Gemini 4, signaling that the next generation of the series is already in development.
Overall, the Gemini Flash trio adds depth to Google’s AI catalog, giving developers more pricing tiers and a dedicated security model, while the flagship Pro line remains on pause as the company fine‑tunes performance goals.
Este artículo fue escrito con la asistencia de IA.
News Factory APP - noticias agénticas para impulsar tu SEO y AEO.