โ† Back to gallery

V8 Model Release Progress and Strategy

Author: GenIArtModel: nanoBanana-ProPublished: 1/21/2026, 9:40:39 PM

Categories

Tags

#AI Art#Image Generation#Personalization#Software Development#V8 Model#Release Strategy#MVP Approach#UI Improvements

Copy prompt

๐Ÿ“…โ›ต Recap Midjourney Office Hours January 21, 2026 In few words: The V8 model is entering its final phase and approaching release with a gradual launch strategy. Major work is underway on prompt management, text, customization, and the interface, with a potential release date in early February. ๐Ÿ‡ซ๐Ÿ‡ท Version franรงaise en commentaire ๐Ÿ”ฝ ๐Ÿ”ถ V8 Image Model Status & Timeline ๐Ÿ”น Primary focus is preparing the V8 image model for launch. ๐Ÿ”น Currently in internal testing (before guides/mods). ๐Ÿ”น Model is in pre-optimization phase, so generation is slower than final. ๐Ÿ”น Many features are still being connected due to: ๐Ÿ”ธ New codebase and new APIs. ๐Ÿ”ธ V8 being significantly different from prior models. ๐Ÿ”น Guides and moderators are expected to start testing this week. ๐Ÿ”น V8 will not release in January. ๐Ÿ”น Earliest likely release window is early February, assuming things go smoothly. ๐Ÿ”น Release timing depends on: ๐Ÿ”ธ Bug fixes. ๐Ÿ”ธ Feature readiness. ๐Ÿ”ธ Optimization duration (could take ~1 week, uncertain). ๐Ÿ”ถ V8 Release Strategy (Rolling / MVP Approach) ๐Ÿ”น Team is debating a non-traditional rollout: ๐Ÿ”ธ Possible early release before full training is complete. ๐Ÿ”ธ Continue training and improving post-launch with frequent updates. ๐Ÿ”น Open question on whether to wait for: ๐Ÿ”ธ Upscaler. ๐Ÿ”ธ Editing tools. ๐Ÿ”น Likely outcome: ๐Ÿ”ธ Launch with a minimal viable feature set, then add features incrementally. ๐Ÿ”น Core challenge is deciding when V8 is โ€œgood enoughโ€ to ship. ๐Ÿ”ถ Prompt Handling & Model Behavior Improvements ๐Ÿ”น Major focus on making V8 work well across all prompting styles: ๐Ÿ”ธ Short prompts. ๐Ÿ”ธ Long prompts. ๐Ÿ”ธ Mixed or unconventional formats. ๐Ÿ”น Current issue: ๐Ÿ”ธ Model is somewhat finicky depending on prompt style. ๐Ÿ”น Goals: ๐Ÿ”ธ Consistent quality regardless of prompt length or structure. ๐Ÿ”ธ Avoid prompt rewriting by the model. ๐Ÿ”ธ Strong out-of-the-box behavior without prompt engineering. ๐Ÿ”ถ Bugs, Testing, and Optimization ๐Ÿ”น At least 1โ€“2 known bugs actively being addressed. ๐Ÿ”น Process: ๐Ÿ”ธ Fix bugs. ๐Ÿ”ธ Retest. ๐Ÿ”ธ Repeat until stable. ๐Ÿ”น Feature hookup and bug iteration happening simultaneously. ๐Ÿ”น Optimization phase still pending after features stabilize. ๐Ÿ”ถ Editing Model & Resolution Plans ๐Ÿ”น Separate edit model is in development alongside V8. ๐Ÿ”น A smaller model is also being explored. ๐Ÿ”น Model supports: ๐Ÿ”ธ Multiple resolutions. ๐Ÿ”ธ Multiple output sizes. ๐Ÿ”น Still undecided: ๐Ÿ”ธ Which resolutions map to which use cases. ๐Ÿ”น Possible future plans: ๐Ÿ”ธ Native 2048 resolution mode. ๐Ÿ”ธ High-resolution training later (not started yet). ๐Ÿ”ถ UI & Style Creator Updates ๐Ÿ”น Ongoing UI improvements across the platform. ๐Ÿ”น Style Creator updates: ๐Ÿ”ธ New scrolling-based interface. ๐Ÿ”ธ Ability to initialize using: โ–ช๏ธ Style references (srefs). โ–ช๏ธ Search terms. ๐Ÿ”น Acknowledgement that: ๐Ÿ”ธ Users still wonโ€™t be fully satisfied. ๐Ÿ”ธ A larger third redesign is planned later. ๐Ÿ”น Short-term plan: ๐Ÿ”ธ Ship incremental improvements soon. ๐Ÿ”ธ Accept tradeoffs rather than blocking progress. ๐Ÿ”ถ Personalization System Overhaul ๐Ÿ”น New scrolling-based personalization interface in development. ๐Ÿ”น Similar interaction style to the updated Style Creator. ๐Ÿ”น Key improvements: ๐Ÿ”ธ No more clicking between two images. ๐Ÿ”ธ Smoother and more intuitive flow. ๐Ÿ”น Current personalization is widely disliked. ๐Ÿ”น New version aims to be: ๐Ÿ”ธ Better overall. ๐Ÿ”ธ Not worse in any major way. ๐Ÿ”น Team feels close to โ€œflipping the switchโ€ to the new system. ๐Ÿ”น Ranking party may still be required before launch: ๐Ÿ”ธ Improves quality. ๐Ÿ”ธ Could delay release slightly. ๐Ÿ”ถ Feature Load & Development Constraints ๐Ÿ”น V8 includes far more than basic text-to-image: ๐Ÿ”ธ Style references. ๐Ÿ”ธ Mood boards. ๐Ÿ”ธ Personalization. ๐Ÿ”ธ Chaos / weird parameters. ๐Ÿ”ธ Style creator. ๐Ÿ”ธ Upscaling. ๐Ÿ”ธ Editing. ๐Ÿ”น Reality: ๐Ÿ”ธ More features than people on the team. ๐Ÿ”น Development effort is spread thin but progressing. ๐Ÿ”ถ Backwards Compatibility & Niji Integration ๐Ÿ”น Personalization and mood boards will be added to Niji G7. ๐Ÿ”น Lack of these features at Niji G7 launch had no strong reasonโ€”just timing. ๐Ÿ”น Focus on: ๐Ÿ”ธ Backwards compatibility. ๐Ÿ”ธ Avoiding unnecessary model splits. ๐Ÿ”ถ Niji v7 Reception & Feedback ๐Ÿ”น Niji v7 response has exceeded expectations. ๐Ÿ”น Strong positive feedback from users. ๐Ÿ”น Noted strengths: ๐Ÿ”ธ Clean, confident line work. ๐Ÿ”ธ Vibrant colors. ๐Ÿ”ธ Improved coherence. ๐Ÿ”ธ Better prompt following. ๐Ÿ”น Text rendering improvements exist, but: ๐Ÿ”ธ Text is not the primary goal of Niji. ๐Ÿ”ธ V8 will handle text significantly better. ๐Ÿ”น Team is actively listening for what users value most. ๐Ÿ”ถ Text Rendering Outlook ๐Ÿ”น V8 will have much stronger text performance than current models. ๐Ÿ”น Users focused on text should expect major gains with V8 rather than Niji. ๐Ÿ”ถ Overall Outlook ๐Ÿ”น V8 is in the final stretch, but still has many moving parts. ๐Ÿ”น Progress is steady, not blocked by a single issue. ๐Ÿ”น Expect: ๐Ÿ”ธ Early February release. ๐Ÿ”ธ Possible rolling updates post-launch. ๐Ÿ”น Team confidence is high; quality is already strong and improving rapidly.

Open on X