AI image service Midjourney launches its new website, with browser-based image generation in the starting blocks.
Midjourney, which has primarily operated through Discord, is looking to bring its full service to the browser. The first version of the new site focuses on better image and prompt search in both user-generated and community-generated images, better image browsing, and faster image viewing with prompts. More organizational features like collections are planned for the coming versions.
The most important new feature is image generation directly in the browser instead of through Discord. It is not known when the image generation in the browser will start, but it should not take much longer: The feature is already visible in the interface, it's just not active yet ("coming soon").
Midjourney's website is a growth opportunity
Midjourney runs very successfully on Discord, but the chat's software complexity is likely to put off potential users. A good browser version is a growth opportunity for the startup.
OpenAI, meanwhile, has just made a strong showing with DALL-E 3. OpenAI's new image AI is not necessarily ahead in terms of pure image quality, at best it is on par, even behind in terms of photorealism. It also offers fewer image editing and generation features and feels more restrictive.
But it follows the prompts so much more accurately, and using it via ChatGPT, just talking to it instead of prompting, is as easy as can be. These are two huge advantages in everyday work.
Midjourney has already announced that version v6 of Midjourney will also offer a significant leap in quality and follow text input better than the current version. It will be released later this year.