Google launches three new Gemini AI models

Google launched three new artificial intelligence models on Tuesday, July 21, 2026, intensifying competition in the AI sector by offering developers faster, cheaper, and more specialized tools for coding, automation, and cybersecurity. The releases include Gemini 3.6 Flash, Gemini 3.5 FlashLite, and the restricted-access Gemini 3.5 Flash Cyber.

Gemini 3.6 Flash is positioned as the primary update, designed for coding, knowledge work, and multimodal tasks. The company says it uses 17 percent fewer output tokens than its predecessor, Gemini 3.5 Flash, leading to greater efficiency and lower cost for developers. It is priced at $1.50 per million input tokens and $7.50 per million output tokens. Google reported performance gains on several benchmarks, including scoring 49 on DeepSWE, up from 37, and 63.9 on MLE Bench, up from 49.7. Early customers like the legal AI firm Harvey and financial research platform Hebbia have used it for document parsing and data analysis.

The second model, Gemini 3.5 FlashLite, is built for high-throughput, low-latency workloads such as translation, summarization, and document processing. It is the most affordable of the new releases, priced at $0.30 per...