May 2026 Summaries

Venice's chat system now employs an "agentic chat" that automates task execution by breaking complex requests into steps, selecting the appropriate tool for each, and delivering results across various media formats such as text, images, and video within a single conversation. This system uses multiple agent models, with the default being Kimi K2.5, and incorporates real-time processing, privacy measures with zero data retention, and a subscription-based model selection that includes free and premium options. Users can engage in multimodal workflows by chaining tasks such as researching, generating images, and creating videos in one prompt, while file uploads are processed locally to ensure privacy. Classic chat remains available, offering specific model selection and specialized modes that are not yet included in the agentic chat, allowing users to switch as needed. Venice is gradually rolling out this new agent-driven chat feature to all users, promising an integrated and efficient way to handle diverse and complex tasks while maintaining control over privacy and model choice.