Google Releases Gemini 3.8 Flash With New Cybersecurity Variant
Table of Contents
Google has released Gemini 3.8 Flash, its newest Flash model for software engineering, agentic workflows and complex reasoning. The launch comes just three weeks after Gemini 3.7 Flash and marks Google’s third Flash release in six weeks.
Google has also introduced Gemini 3.8 Flash Cyber, a specialized variant designed to help trusted defenders identify software vulnerabilities and develop patches. Access to the cybersecurity model is being provided through Google’s new Fairwind Program.
Gemini 3.8 Flash Focuses on Coding and AI Agents
Google describes Gemini 3.8 Flash as its most intelligent workhorse model, with improvements in software engineering, agentic tasks and multi-step reasoning.
The model can perform additional reasoning steps and use tools repeatedly when handling complex tasks. This makes it particularly suited to software development, where AI agents may need to inspect code, plan changes, use development tools and evaluate their work.
Google reports stronger results across several software engineering and professional benchmarks. These are Google’s reported results and should not be treated as independent proof that the model performs better in every workload.
Gemini 3.8 Flash May Use More Tokens
The model’s stronger reasoning can also increase its overall token usage.
Google says Gemini 3.8 Flash may perform extra reasoning and make repeated tool calls on difficult tasks. As a result, a task could cost more overall even when the per-token price remains unchanged.
Developers focused on efficiency can adjust reasoning effort or continue using Gemini 3.7 Flash for workloads where lower computational usage is more important.
Gemini 3.8 Flash Pricing
Google is offering the model at an introductory price of $0.75 per million input tokens and $3.75 per million output tokens through December 31, 2026.
From January 1, 2027, the planned prices will increase to $1.50 per million input tokens and $7.50 per million output tokens.
| Details | Information |
| Main focus | Coding, agents and reasoning |
| Input price until Dec. 31, 2026 | $0.75 per million tokens |
| Output price until Dec. 31, 2026 | $3.75 per million tokens |
| Input price from Jan. 1, 2027 | $1.50 per million tokens |
| Output price from Jan. 1, 2027 | $7.50 per million tokens |
| Previous Flash model | Gemini 3.7 Flash |
Gemini 3.8 Flash Cyber Targets Vulnerabilities
Gemini 3.8 Flash Cyber is designed specifically for cybersecurity defense.
Working with Google’s CodeMender agent harness, the model can help identify vulnerabilities, develop fixes and validate patches. Google says the system can autonomously find and address vulnerabilities within controlled environments.
Google has reported strong performance on its cybersecurity evaluations, including CyberGym and an internal test covering codebases written in 20 programming languages.
These results come from Google’s own testing, so real-world performance may vary depending on the software, security environment and implementation.
Google Introduces the Fairwind Program
Google is making the cybersecurity model available through its Fairwind Program, a limited-access initiative for trusted defenders.
The program is aimed at governments, national cyber authorities, critical infrastructure operators and software maintainers. Google says more than 650 partners are participating globally.
The goal is to help organizations find and fix vulnerabilities before they can be exploited, while keeping the more powerful cybersecurity capabilities within controlled environments.
Gemini 3.8 Flash Includes Safety Measures
The launch also highlights the safety challenges of advanced AI agents.
Google says Gemini 3.8 Flash includes safeguards covering cybersecurity and other high-risk areas. The Cyber variant uses a different cybersecurity safety configuration because it is intended for trusted defenders.
Google has also reported improvements in protection against prompt-injection attacks, which can be particularly important for AI agents that interact with external tools, code and data.
Where Gemini 3.8 Flash Is Available
Developers can access Gemini 3.8 Flash through Google’s AI development platforms, while enterprise customers can use it through Gemini Enterprise.
Google AI Pro and Ultra subscribers can also access the model through the Gemini app, AI Mode in Google Search and Gemini in Google Sheets.
The Cyber version has more restricted availability through the Fairwind Program.
Why the Rapid Release Pace Matters
The launch continues Google’s fast development cycle for its Flash models.
Arriving only three weeks after Gemini 3.7 Flash, the release shows how quickly AI companies are developing models for coding and agentic workflows.
For developers, frequent releases can bring new capabilities, but they also make it important to evaluate performance, token usage and pricing before moving production workloads to a newer model.
Frequently Asked Questions
What is Gemini 3.8 Flash?
Gemini 3.8 Flash is Google’s latest Flash model for software engineering, agentic workflows and complex reasoning.
How much does Gemini 3.8 Flash cost?
The introductory API price is $0.75 per million input tokens and $3.75 per million output tokens through December 31, 2026.
What is Gemini 3.8 Flash Cyber?
It is a specialized cybersecurity model designed to help trusted defenders discover vulnerabilities and develop and validate software fixes.
What is the Fairwind Program?
Fairwind is Google’s limited-access program that provides advanced cybersecurity AI capabilities to selected governments, critical infrastructure operators and software maintainers.
Is Gemini 3.8 Flash Cyber publicly available?
No. Access is restricted to selected trusted defenders through the Fairwind Program.
Conclusion
The release of Gemini 3.8 Flash marks another rapid development in Google’s AI model lineup, with a focus on coding, reasoning and agentic workflows.
Alongside it, Gemini 3.8 Flash Cyber brings these capabilities into cybersecurity, helping trusted organizations identify vulnerabilities and develop software fixes.
The Fairwind Program provides a controlled way to offer these advanced capabilities to governments, critical infrastructure operators and software maintainers. Together, the two launches show Google’s growing focus on AI systems that can reason through complex tasks and take useful actions rather than simply generate responses.