July 2025 Summaries

Venice has introduced a Social Feed feature, transforming its platform from a private AI generation tool into a community space where users can share AI-generated images, discover creative prompts, and interact with other AI artists. This new feature allows users to showcase their creations on a public social feed, follow their favorite creators, and engage with the community by upvoting or saving posts. Pro users have the option to share their prompts, providing a learning opportunity for others, while maintaining control over what they share. The platform encourages users to actively participate by engaging with content, experimenting with shared prompts, and curating a personalized feed through following features, ultimately fostering a collaborative and inspiring environment for AI art enthusiasts.

Venice's uncensored AI model, Dolphin Mistral 24B Venice Edition, is now accessible for free via OpenRouter's API, providing developers with a powerful, private, and unrestricted AI experience. This partnership allows developers to access multiple AI models through a single API, facilitating seamless integration without the need to rebuild applications, and ensures that conversations remain private with no content restrictions. Venice's model offers a minimal censorship refusal rate of 2.2%, significantly lower than mainstream alternatives like Claude and GPT-4o-mini. The collaboration underscores a commitment to developer autonomy and creative freedom, challenging the restrictive content policies often imposed by mainstream AI platforms. OpenRouter's CEO, Alex Atallah, and Venice's founder, Erik Voorhees, emphasize the partnership's role in expanding access to scalable, sustainable, and transparent AI, providing a valuable alternative for developers seeking privacy and freedom in their AI tools.

Venice Pro offers users advanced image editing capabilities through intuitive, text-based prompts, eliminating the need for mastering complex software. Users can upload or generate images and describe the desired changes in plain language, with the system automatically applying modifications while maintaining image quality. The platform supports four main types of editing: removing and replacing objects, transforming artistic styles, adjusting colors and lighting, and providing detailed prompts for better editing results. Techniques for effective use include using precise language, starting with simple edits, specifying unchanged elements, breaking complex changes into steps, directly naming subjects, quoting exact text, controlling composition, and choosing verbs carefully. Venice Pro's tools enable users to make professional-level adjustments effortlessly, transforming their creative workflows through these accessible features.

Venice has rebranded its measurement of API compute capacity by renaming Venice Compute Units (VCU) to Diem and adjusting their denomination, aiming to simplify the system for users. Under this new naming system, one Diem equates to 10 former VCUs and represents $1 of daily API credit, as opposed to the previous $0.10 per VCU, effectively making each Diem 10 times more powerful than a VCU. While the total capacity numbers now appear smaller due to this change—yesterday's 181,480 VCU capacity is now 18,148 Diem—the actual compute power and resource allocation remain unchanged, ensuring users still receive the same level of API capacity. Users can view their updated Diem allocation and track their daily API capacity on the venice.ai/token dashboard.