Gemini 3.5 Flash Cyber Spots Vulnerabilities Across Google's Internal Codebases
xennygrimmato_ · x · 2026-07-21
Benchmarks only tell part of the story. Paired with the CodeMender agent, Gemini 3.5 Flash Cyber is already finding and fixing vulnerabilities in Google’s internal codebases, including Chrome, Android, Cloud, Ads, and YouTube.
The model was evaluated on the CyberGym benchmark against hundreds of real-world software vulnerabilities. Leveraging its low cost, CodeMender was configured to call 3.5 Flash Cyber up to five times for a single final report, allowing the overall agent to achieve highly competitive performance.
Related event: Google Launches Three New Gemini Models, Announces Gemini 4 Pre-training(157 posts)→
More from coding & agent
- Cognition's SWE-2 uses a KKT duality argument in RL to shift the effort Pareto curve — YouJiacheng · 2026-09-11
- First-ever Three.js Conference lands in Paris, with a panel on AI-shortened design workflows — OdinLovis · 2026-09-11
- Data engineering, not agent frameworks, is the real bottleneck for enterprise AI agents — dhruv2038 · 2026-09-11
- RTK Terminal Compression Cuts Tokens but Leaves Your AI Coding Bill Unchanged — Bartaseth · 2026-09-11
- GPT-6 Astra beats Factorio with enemies in 44 in-game hours at ~$4,500 API cost — liminal_bardo · 2026-09-11
- Investment Analyst Asks How to Build a Claude-Based Diligence Agent Stack — Careless_Tie2286 · 2026-09-11