Google has announced the release of three new Gemini models—Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber—designed to enhance efficiency, latency, and reliability for agentic workflows. Gemini 3.6 Flash is the company's new workhorse model, offering improved coding and multimodal performance while reducing output token usage by 17% compared to its predecessor. Pricing for 3.6 Flash is set at $1.50 per 1 million input tokens and $7.50 per 1 million output tokens. Benchmark improvements for the model include 49% versus 37% on DeepSWE, 63.9% versus 49.7% on MLE Bench, and 83.0% versus 78.4% on OSWorld-Verified.

Gemini 3.5 Flash-Lite is optimized for high-throughput, low-latency tasks such as agentic search. It achieves speeds of 350 output tokens per second, with pricing at $0.30 per 1 million input tokens and $2.50 per 1 million output tokens. The model demonstrates significant gains over previous versions, achieving 54.2% on SWE-Bench Pro and 74.0% on OSWorld-Verified. It is now rolling out in Google Search and is available via the Gemini API, Google AI Studio, and Android Studio. Computer use capabilities are now provided as a built-in client-side tool across these surfaces.

Gemini 3.5 Flash Cyber, a model specialized for cybersecurity, is designed to detect and patch vulnerabilities. It is fine-tuned to function within CodeMender, an agentic infrastructure that utilizes multiple agents to generate security reports. Due to the potential for dual-use misuse, Google is limiting access to 3.5 Flash Cyber to governments and trusted partners through a pilot program. Additionally, Google confirmed that Gemini 3.5 Pro is currently in testing and that pre-training for the future Gemini 4 model has commenced.

Google has announced the release of three new Gemini models—Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber—designed to enhance efficiency, latency, and reliability for agentic workflows. Gemini 3.6 Flash is the company's new workhorse model, offering improved coding and multimodal performance while reducing output token usage by 17% compared to its predecessor. Pricing for 3.6 Flash is set at $1.50 per 1 million input tokens and $7.50 per 1 million output tokens. Benchmark improvements for the model include 49% versus 37% on DeepSWE, 63.9% versus 49.7% on MLE Bench, and 83.0% versus 78.4% on OSWorld-Verified.

Gemini 3.5 Flash-Lite is optimized for high-throughput, low-latency tasks such as agentic search. It achieves speeds of 350 output tokens per second, with pricing at $0.30 per 1 million input tokens and $2.50 per 1 million output tokens. The model demonstrates significant gains over previous versions, achieving 54.2% on SWE-Bench Pro and 74.0% on OSWorld-Verified. It is now rolling out in Google Search and is available via the Gemini API, Google AI Studio, and Android Studio. Computer use capabilities are now provided as a built-in client-side tool across these surfaces.

Gemini 3.5 Flash Cyber, a model specialized for cybersecurity, is designed to detect and patch vulnerabilities. It is fine-tuned to function within CodeMender, an agentic infrastructure that utilizes multiple agents to generate security reports. Due to the potential for dual-use misuse, Google is limiting access to 3.5 Flash Cyber to governments and trusted partners through a pilot program. Additionally, Google confirmed that Gemini 3.5 Pro is currently in testing and that pre-training for the future Gemini 4 model has commenced.