Gemini 3.8 Flash is available at an introductory price of $0.75 per million input tokens and $3.75 per million output tokens through the end of the year. The new model targets software engineering and multi-step reasoning, while a specialized Gemini 3.8 Flash Cyber variant offers automated patching and vulnerability detection for members of the Fairwind Program. Access to the Cyber version is limited to trusted defenders to secure code and find security flaws.
The launch is the third Flash series update in six weeks and includes a 1 million token context window with support for text, image, video, and speech inputs. In benchmarks, 3.8 Flash scored 62.3% on the Vals Index and 59 on the Artificial Analysis Intelligence Index, a 3 point increase over version 3.7. High-reasoning output speeds average 300 tokens per second, and standard pricing will increase to $1.50 and $7.50 per million tokens for input and output respectively in 2027.
Developers can access the model through Google AI Studio, Android Studio, and the Gemini API. Enterprise and consumer availability extends to the Gemini App for Pro and Ultra subscribers, AI Mode in Search, and Google Sheets.
Key sources
- SOURCE@googledeepmind“frontier-level vulnerability detection and automated patching”x.com
- SOURCE@google“$0.75/M input tokens and $3.75/M output tokens”x.com
- SUPPORT@google“Apply for access to Gemini 3.8 Flash Cyber through our Fairwind Program”x.com
- SOURCE@koraykv“sits on the Pareto Frontier on CWE-Bench for patching”x.com
- SUPPORT@sundarpichai“3rd Flash release in just 6 wks”x.com
- SUPPORT@_philschmid“verifies its work more often”x.com
- SUPPORT@artificialanlys“scores 59 on the Artificial Analysis Intelligence Index”x.com
- SUPPORT@valsai“Gemini 3.8 Flash scores 62.3% on the Vals Index”x.com