Google has announced Gemini 3.8 Flash, a new model that claims to be its best for reasoning and coding tasks, while also demonstrating strong visual capabilities. A specialized version, Gemini 3.8 Flash Cyber, has reportedly outperformed GPT models in identifying software vulnerabilities. AI
IMPACT Sets a new benchmark for reasoning and coding, potentially influencing future model development and competition.
RANK_REASON Frontier-lab model release with system card [lever_c_demoted from frontier_release: ic=1 ai=1.0]
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →