Google’s Gemini apps are seeing rapid upgrades across models, creativity, and workflow tools. The latest release introduces an enhanced Gemini 2.5 Flash that produces more organized responses with headers, lists, and tables, plus stronger image understanding (2025.09.25). Custom Gems can now be shared through familiar Google Drive style permissions, making it easier to distribute reusable, purpose-built assistants across friends, family, and teams (2025.09.18). Recent additions also include chat history search on web and mobile (2025.08.21) and a Temporary Chat mode that does not save to history or train models (2025.08.13).
Creative features continue to expand. A new image generation and editing model improves instruction following, multi image blending, and consistent selfie rendering (2025.08.26). Imagen 4 raises text to image quality with sharper details and better typography for use cases like posters and invites (2025.05.20). Video generation advances include 8 second videos with sound via Veo 3 for Google AI Ultra subscribers (2025.05.20) and earlier Veo 2 access for Gemini Advanced subscribers (2025.04.22). Canvas adds a Create menu for generating web pages, infographics, quizzes, Audio Overviews, and even working prototypes with auto error resolution, while vibe coding apps gain persistence, multi user data sharing, and home screen shortcuts (2025.05.20). Audio Overview also brings podcast style conversations derived from user files and reports (2025.03.18).
Learning and productivity are a major focus. Users over 18 can generate unlimited custom quizzes, flashcards, and study guides (2025.08.06), while Gemini now weaves high quality images, diagrams, and YouTube videos into explanations and offers Guided Learning for step by step understanding (2025.08.06). Scheduled actions let subscribers plan recurring or one off tasks from summaries to local event suggestions (2025.06.19). A Productivity Planner Gem connects Gmail, Calendar, and Drive to surface priorities and action items for eligible Pro, Ultra, and qualifying Workspace plans (2025.07.15). Document upload is available to all users with a 32K context window, while Gemini Advanced retains a 1 million token context window and higher limits (2025.02.20). Deep Research now runs on Gemini 2.0 Flash Thinking, is available to try at no cost, and supports user file and image sources with Canvas integrations (2025.05.20 and 2025.03.13).
Model upgrades are frequent. Google AI Ultra subscribers get early access to Gemini 2.5 Deep Think for longer, parallel reasoning on complex tasks (2025.08.01). Gemini 2.5 Pro (experimental) expanded to all users with Canvas and later received a coding boost (2025.03.29 and 2025.05.06). The service added 2.5 Flash (experimental) and upgraded Gemini 2.0 Flash Thinking for improved advanced reasoning and app connected task execution (2025.04.17 and 2025.03.13). Gemini itself moved to 2.0 Flash for faster, stronger baseline performance (2025.01.30).
Platform and access are broadening. Gemini in Chrome begins rolling out to Google AI Pro or Ultra subscribers in the U.S. on Windows and Mac OS, bringing page aware assistance without tab switching (2025.05.20). Google introduced Google AI Ultra for top tier access to models like 2.5 Pro, Deep Research, Veo 3, and a 1 million token context window, and renamed Gemini Advanced under the Google AI Pro subscription while retaining core benefits (2025.05.20). Students over 18 in Indonesia, Japan, the UK, and Brazil get a free upgrade through July 2026, including NotebookLM and 2 TB storage (2025.05.20). New Android and service integrations include Spotify, Phone, Messages, WhatsApp, and Utilities extensions, plus an Android overlay for contextual help (2025.03.03 and 2024.08.26).
