For Developers/Models/Gemini 3.5 Flash Cyber
Closed SourcePreviewGoogleReleased 2026-07

Gemini 3.5 Flash Cyber

Cost-efficient cybersecurity-focused model for vulnerability detection and patching

Context window

Not announced

Input / 1M tokens

Not announced

Output / 1M tokens

Not announced

Provider

Google

Data verified 2026-07-22

Gemini 3.5 Flash Cyber is a specialized model built on top of Gemini 3.5 Flash and fine-tuned for finding and fixing cybersecurity vulnerabilities. Designed to detect, validate, and patch code security issues at scale at a lower price per token than larger models. Deployed through CodeMender with multi-agent orchestration for comprehensive vulnerability scanning.

Capability index

Relative estimates (0-100) to place this model against its peers, grounded in published benchmarks.

Coding
9
Reasoning
7
Math
5
Multimodal
0
Long context
7
Speed
8
Cost efficiency
9

How to access it

Exclusively available to governments and trusted partners via CodeMender as part of a limited-access pilot program.

Strengths

  • Lower cost per token than large cybersecurity models
  • Discovered 55 unique vulnerabilities in V8 (vs 47 for standard 3.5 Flash)
  • Efficient multi-agent architecture for scanning large codebases
  • Proven real-world performance finding RCE and memory-corruption bugs

Best for developers who...

Automated vulnerability detection at scaleCost-sensitive security scanning workflowsOrganizations needing efficient code security analysis

When to choose it (and when not to)

Reach for Gemini 3.5 Flash Cyber when...

  • When you need continuous vulnerability scanning across large codebases
  • For organizations that have access through the limited pilot program

Look elsewhere if...

  • If you need public self-serve API access (currently restricted to governments and partners)
  • For use cases requiring broader model availability

How to use it

Quickstart

N/A
Access restricted to pilot participants via CodeMender

Public API not available; limited-access pilot program only

API model id: Not announced

Benchmarks

BenchmarkScoreNotes
V8 JavaScript Engine (Chrome)55 unique vulnerabilities foundOutperformed 3.5 Flash (47) and Claude Opus 4.6 (36)
CyberGymCompetitive frontier-level performanceMulti-agent CodeMender configuration
Big Sleep (Complex Codebase Security)Significant uplift vs 3.5 FlashTested on real-world vulnerabilities in Chrome and Safari

Source: Google Official Blog

Compare Gemini 3.5 Flash Cyber

Compare Gemini 3.5 Flash Cyber with any other model

Build a comparison →
All model comparisons →

Learn the concepts