Google DeepMind Introduces Gemini 3.5 Flash Cyber for Vulnerability Detection

Today, we’re expanding our longtime efforts to better prepare defenders by introducing Gemini 3.5 Flash Cyber, our lightweight cybersecurity model built on top of 3.5 Flash and fine-tuned to find, validate, and patch vulnerabilities quickly and efficiently, making it more effective at these tasks than Gemini’s mainline Flash models. Flash’s performance and efficiency makes it an ideal foundation for our cybersecurity model efforts. By building on top of Flash, 3.5 Flash Cyber offers a cost-efficient and highly capable alternative to large, costly cybersecurity models. [...] Given the dual-use nature of this technology, we have taken an intentional approach to how we deploy 3.5 Flash Cyber. As part of a limited-access pilot program, 3.5 Flash Cyber will be exclusively available to governments and trusted partners via CodeMender soon, expanding over time. This will give frontline defenders a head start in finding and fixing critical vulnerabilities before they can be exploited, while mi

In this article, we’ll explore: In the second part of this series, we’ll examine benchmark results from CyberGym, Big Sleep, Chrome’s production pipeline, and compare Flash Cyber against Claude Opus 4.6 and other leading AI models. ## What Is Gemini 3.5 Flash Cyber? Gemini 3.5 Flash Cyber is a specialized cybersecurity language model built on top of Gemini 3.5 Flash and fine-tuned specifically for vulnerability discovery, validation, and remediation. Unlike Gemini 3.6 Flash, which serves as a general-purpose reasoning and coding model, Flash Cyber has a much narrower objective: Help defenders automatically discover, verify, and fix software vulnerabilities before attackers can exploit them. [...] To address this problem, Google introduced Gemini 3.5 Flash Cyber, a specialized cybersecurity model announced alongside Gemini 3.6 Flash and Gemini 3.5 Flash-Lite. Unlike traditional general-purpose AI assistants, Flash Cyber is purpose-built to find, validate, and patch software vulnera

googledeepmind's profile picture googledeepmind Finding security flaws in massive software projects means searching through millions of lines of code. To help defenders keep up, we built Gemini 3.5 Flash Cyber. It’s a fast, lightweight AI model made specifically to scan large codebases, spot subtle bugs, and help teams fix vulnerabilities before they can be exploited. 🛡️ mohammed24208's profile picture mohammed24208 We want pro Reply nowwebuildshow's profile picture nowwebuildshow A model built to catch bugs before hackers do, that's the right side of the AI race to be on. Reply dariush2000b's profile picture dariush2000b 🔥🔥🔥 Reply dariush2000b's profile picture dariush2000b ❤️❤️❤️ Reply howisaichangingscience's profile picture howisaichangingscience

1,641,007 followers Gemini 3.5 Flash Cyber is our specialized, lightweight model built to help security teams spot and patch vulnerabilities before they can be exploited. By combining speed and efficiency, 3.5 Flash Cyber lets defense specialists check significantly more code paths. In testing on our own codebases – including Google Chrome and Android – it consistently caught complex, unique vulnerabilities that standard models missed. To ensure this model is deployed responsibly, we’re starting with a limited-access pilot for governments and trusted partners. → No alternative text description for this image Cybernorse, graphic [...] View organization page for Google DeepMind 1,641,007 followers Gemini 3.5 Flash Cyber is our specialized, lightweight model built to help security teams spot and patch vulnerabilities before they can be exploited. By combining speed and efficiency, 3.5 Flash Cyber lets defense specialists check significantly more code paths. In testing on our own cod

#### World's Premier Cyber Security Portal Google DeepMind has introduced Gemini 3.5 Flash Cyber, a specialised artificial intelligence model designed to discover, validate and help remediate security vulnerabilities across large software codebases. Built on Gemini 3.5 Flash and fine-tuned specifically for cybersecurity work, the lightweight model is intended to give defensive teams a faster and more economical alternative to using large frontier models for every stage of vulnerability research. [...] Article content ### A lightweight model designed to search more code Gemini 3.5 Flash Cyber is based on the idea that effective vulnerability discovery does not always depend on making a single request to the largest and most expensive model available. Finding a deeply embedded software flaw often requires examining an enormous number of possible code paths, tracing how untrusted data moves through a program, testing assumptions and repeatedly adjusting the analysis as new evidence e

Log inSign up ## Post user avatar Google DeepMind @GoogleDeepMind Gemini 3.5 Flash Cyber is our specialized, lightweight model built to help security teams spot and patch vulnerabilities before they can be exploited. 🧵 GIF 3:57 PM · Jul 23, 2026124KViews By combining speed and efficiency, 3.5 Flash Cyber lets defense specialists check significantly more code paths. In testing on our own codebases – including @googlechrome and @Android – it consistently caught complex, unique vulnerabilities that standard models missed. 25K user avatar Google DeepMind @GoogleDeepMind Jul 23 To ensure this model is deployed responsibly, we’re starting with a limited-access pilot for governments and trusted partners. → [...] Introducing Gemini 3.5 Flash CyberFrom deepmind.google 21K user avatar Ivana @ivanainai Jul 23 Gemini hasn’t improved, but at least Google moved the buttons again 1.3K # Join the conversation Read 97 more replies