01 What happened
Google has expanded its model lineup with Gemini 3.8 Flash, designed for complex software engineering and multi-step reasoning, and Gemini 3.8 Flash Cyber, which focuses on automated security.
02 Key details
- Gemini 3.8 Flash improves upon its predecessor, achieving a 54.9% score on the HLE-Verified benchmark for multi-step reasoning.
- Gemini 3.8 Flash Cyber provides automated vulnerability detection with a success rate exceeding 70% across 20 programming languages.
- The Cyber model reaches a pass@1 of 47.2% on the CWE-Bench benchmark for automated code patching.
- The 3.8 model series includes enhanced robustness against prompt injection attacks, as verified by Gray Swan testing.
03 Why it matters
These models provide enterprises and developers with cost-efficient tools for autonomous agent workflows and automated software security patching.
04 Who it matters to
Software developers, cybersecurity specialists and AI integration specialists.
Original sourceGoogle