Google Releases Gemini 3.8 Flash — Third Flash Model in Six Weeks

Google has released Gemini 3.8 Flash, its third Flash model in six weeks. At the same introductory price, the model performs more reasoning steps and consumes more tokens on complex tasks, while a closed cyber model, Gemini 3.8 Flash Cyber, is also introduced for finding and patching vulnerabilities.

Compiled by Sergey KostenchukPublished 2026-09-02Updated 2026-09-18

2026-09-02 Research Google

🤖 Google has released Gemini 3.8 Flash, its third Flash model in six weeks.

The introductory price remains the same: $0.75/$3.75 per million input/output tokens until December 31, 2026, then $1.50/$7.50. On complex tasks, the model performs more reasoning steps and consumes more tokens. For budget-conscious users, there are reduced effort levels and support for 3.7 Flash. On HLE-Verified, it scores 54.9%, with a context window of up to 1 million tokens.

🌍 Quality is improving at the same rate, but the bill for complex tasks is rising due to token consumption: the market is shifting from "price per token" to "price per solved task."

👤 The model is already available in the Gemini API, AI Studio, Antigravity, Android Studio, and the Gemini app. Calculate your budget based on token consumption telemetry. The Cyber version for vulnerability discovery is only available through Fairwind.

Source 1: https://blog.google/innovation-and-ai/models-and-research/gemini-models/3-8-flash-and-3-8-flash-cyber/ Source 2: https://deepmind.google/models/model-cards/gemini-3-8-flash/

Sources