Gemini 2.5 Flash-Lite, previously in preview, is now stable and generally available. This cost-efficient model is ~1.5x faster than 2.0 Flash-Lite and 2.0 Flash, offers high quality, and includes 2.5 family features like a 1 million-token context window and multimodality.

Gemini's advanced capability for conversational image segmentation allows intuitive interaction with visual data by understanding complex phrases, conditional logic, and abstract concepts, streamlining developer experience and opening doors for new applications in media editing, safety monitoring, and damage assessment.

Veo 3, Google’s latest AI video generation model, is now available in paid preview via the Gemini API and Google AI Studio. Unveiled at Google I/O 2025, Veo 3 can generate both video and synchronized audio, including dialogue, background sounds, and even animal noises. This model delivers realistic visuals, natural lighting, and physics, with accurate lip syncing and sound that matches on-screen action.

Updates in Firebase Studio include new Agent modes, foundational support for the Model Context Protocol (MCP), and Gemini CLI integration, all designed to redefine AI-assisted development allow developers to create full-stack applications from a single prompt and integrate powerful AI capabilities directly into their workflow.

The `logprobs` feature has been officially introduced in the Gemini API on Vertex AI, provides insight into the model's decision-making by showing probability scores for chosen and alternative tokens. This step-by-step guide will walk you through how to enable and interpret this feature and apply it to powerful use cases such as confident classification, dynamic autocomplete, and quantitative RAG evaluation.