While a viral phrase overstates where Midjourney actually is right now, the company's quiet V8 Alpha release and its maturing video pipeline suggest the underlying reality is almost as significant as the hype.
"Image 2.0 cooked here" has been circulating on X and across AI subreddits this week, attached to outputs from Midjourney's latest model generation. The phrase captures something real about where the platform has arrived, even if the specific trigger being cited online contains some inaccuracies worth clearing up. Midjourney V7 did not drop in April 2026. It launched on April 3, 2025, and became the platform's default model in June of the same year. What actually happened in March 2026 was the release of a V8 Alpha , quieter, less announced, distributed first to the Discord community , and it is that model, alongside Midjourney's evolving video capabilities, that is driving the current conversation.
The V8 Alpha is visibly different from V7 in the areas that matter most to professional users: semantic understanding of complex multi-subject prompts, text rendering accuracy, and spatial coherence in scenes with multiple interacting elements. Independent tests run by the AI Video Bootcamp team across 30 standardized prompts found V7 already outperformed V6 on photorealism in 23 of those cases, with skin textures, fabric detail, and shadow rendering all showing measurable gains. V8 Alpha advances that further, with community feedback consistently flagging improved handling of prompts that previously caused element merging , the persistent problem where a scene described with three or more distinct objects would collapse into visual noise. A new Edit model is also in development alongside V8, bringing inpainting, outpainting, and multi-reference support to a platform that has historically required third-party tools for those workflows. V9 is confirmed as in progress, with Niji 8 and a Video V2 model also publicly referenced on Midjourney's development roadmap.
The video situation is where Midjourney's 2026 trajectory becomes genuinely competitive. The 2026 Web Editor added an "Animate" button that uses the Video V1 model to convert static generations into smooth high-definition clips. Users on the updated interface report generating 21-second videos from single images, a meaningful step up from the 10-second outputs that defined the initial video launch. Video V2 is in development and expected to deliver substantially improved motion consistency and longer clip lengths, bringing Midjourney into direct competition with Runway Gen-4, Kling 2.0, and Google's Veo 2 , all of which have raised the baseline for what professional AI video looks like in Q2 2026.
The Platform's Structural Advantages
What makes Midjourney's position interesting is what it has preserved while competitors scaled. David Holz has run the company without major institutional investment, and Midjourney remains profitable on subscription revenue alone, something almost no comparably capable AI lab can claim. As of April 2026, the platform reports 22 million registered members. Pricing has stayed unchanged at $10 to $120 per month across tiers, and the shift from Discord-only to a full web interface has removed the primary friction point that kept less technical users off the platform. The web editor's "Semantic Control" adjacent features , which allow users to reference specific elements across generations using Omni Reference and style references , are now stable enough for production workflows, not just experimentation.
The competitive framing is worth being precise about. Midjourney's strength has always been aesthetic quality and artistic coherence rather than photorealistic simulation at scale. OpenAI's Sora, xAI's Grok Imagine, and Runway have all pushed further on video realism and temporal consistency. What Midjourney has that most competitors do not is a trained user community with deep prompt literacy, a personalization system that learns individual aesthetic preferences over time, and a price-to-quality ratio that remains difficult to beat for still image work. The question V8 and Video V2 will answer is whether those structural advantages extend into video, or whether that market consolidates differently from the image generation market Midjourney already leads.
Why the Viral Phrase Landed
"Cooked" as slang means decisively outclassed. The phrase being applied to Midjourney's outputs , rather than to Midjourney by its competitors , reflects genuine user sentiment about where image quality has arrived. An independent creator in April 2026 can generate product photography, concept art, editorial illustration, and now short video content from a $10 monthly subscription, at a quality level that would have required a professional studio budget three years ago. That shift is not primarily about any single model release. It is the cumulative result of Midjourney's iterative release cadence: V6 to V7 to V8 Alpha in roughly eighteen months, each generation compressing the gap between what a user imagines and what appears on screen. Whether the video pipeline catches up to the image pipeline in the same timeframe is the story worth watching for the rest of 2026.
Also read: Elon Musk's viral Grok video is a stress test for how we handle AI-generated reality • Isomorphic Labs is putting AI-designed drugs into humans and the results will define a decade • Anthropic's Mythos is a real threat to crypto infrastructure, just not in the way the panic suggests
