🚀 Midjourney Office Hours Highlights: January 8th

As Midjourney resumes its weekly office hours following a holiday break, the focus is firmly set on the upcoming V7 model and other significant projects. These sessions serve as a platform to share ongoing work, gather community feedback, and address any questions from users. The anticipation is palpable as the team gears up to unveil a series of groundbreaking advancements.

For a quick overview, head straight to the TL;DR; section otherwise keep on reading.

Extended Recap

🛠️ V7 Development: A New Era

The V7 model represents a complete overhaul of its predecessor, V6. This ambitious project involves a new architecture, different data sets, and revised language processing approaches. The primary goal is to enhance prompt understanding, coherence, and flexible styling. Currently, the model is undergoing rigorous internal testing and bug fixing, with a possible timeline ranging from a week to a month for testing, followed by a two-week final tuning phase. An alpha period, often with limited features, is expected to follow. The delay in V7's release is attributed to the extensive reworking of almost every part of the system. Future expansions are anticipated, including draft modes, large-batch modes, and specialized offshoots that vary in speed and quality.

🎥 Video Feature Explorations: Innovating the Experience

Midjourney is exploring multiple potential solutions for video features. These include an in-house video model, possible partnerships with external models, and exploration of open-source baselines. The key challenge lies in making the video experience enjoyable without being overly time-consuming or costly. The team is working to balance speed, cost, and output quality to ensure a seamless user experience.

📈 Larger Batch Mode & Fast Modes: Enhancing User Interaction

The development of a "bigger batch mode" on the website is underway, allowing users to iterate over larger sets of images, marking their preferences. The goal is to implement both super-fast generation, akin to a "real-time" mode, and high-capacity modes for large-scale creation. Personalization improvements are also in progress, potentially reducing setup time by 30%.

🛌 Relaxathon and Community Usage: Unlimited Creativity

Midjourney has introduced temporary "Relax Mode" servers, offering users essentially unlimited image generation at no extra cost, albeit typically at slower speeds. However, due to extra server capacity, the current speeds are not slower. Users are encouraged to take advantage of this opportunity before servers are reassigned to V7 and video.

🖼️ Mood Boards & Editor Features: Crafting Unique Styles

Recent releases include mood boards and multiple personalization profiles. Future improvements may combine mood boards with S-Ref (style/reference) or "character reference" capabilities. There is potential for an Explore page dedicated to mood boards, though this is still under debate. Editor enhancements, including re-texturing and potential expansions of features, remain a work in progress.

🎨 Personalization System: Tailoring the Experience

Midjourney offers two main methods for personalization: quick "image pair" preference ranking (profiles) and mood boards for specific style references. Plans are in place to encourage users who have never tried personalization, as feedback indicates that around 85% of people prefer images generated using personalization. The current focus is on making the personalization process faster rather than adding complexity.

🔔 Conclusion: Embrace the Innovations

As the team concentrates on finalizing V7 and testing video features, users are encouraged to make generous use of Relax Mode while extra servers are available. Experimenting with mood boards, personalization, and any alpha/beta features as they roll out is highly recommended. Ongoing weekly office hours will continue to provide updates and Q&A opportunities, ensuring that the community stays informed as testing and feature rollouts progress.

TL;DR

🚀 V7 Development: A complete overhaul of V6 with new architecture and data sets, focusing on prompt understanding and flexible styling. Testing is ongoing, with future expansions planned.

🎥 Video Features: Exploring in-house and external solutions to make video experiences enjoyable and cost-effective.

📈 Batch & Fast Modes: Developing larger batch modes and super-fast generation options, with personalization improvements underway.

🛌 Relaxathon: Unlimited image generation available temporarily at no extra cost, encouraging users to explore before resources shift.

🖼️ Mood Boards: Recent releases include mood boards, with future enhancements planned for style and reference capabilities.

🎨 Personalization: Focus on quick preference ranking and mood boards, aiming to streamline the process for users.

Stay tuned for more updates as Midjourney continues to innovate and enhance its platform!

If you want to support me, feel free to buy me a coffee ☕️ 

If you’re not subscribed yet to my newsletter Imagine Weekly, I’d be thrilled to welcome you on board!