[ad_1]
Midjourney is once again allowing image creation for non-subscribers. This is a temporary relaxation, warns the startup behind the generative AI system.
It was over a year ago. In the spring of 2023, Midjourney decided to no longer allow free accounts to create images with its generative artificial intelligence system. This decision was made at the time shortly after the release of version 5 of its AI. But today, the young American startup is softening its position.
In a tweet published on its account on August 21, 2024, Midjourney announced that it is once again possible to create images for free. This opening accompanies the global release of the platform’s website, which offers the possibility of generating visuals without going through the Discord application. In fact, the web experience is intended to be more accessible.
Be careful though: Midjourney specifies that this relaunch is, as things stand, temporary. The company does not detail the contours of its operation, which aims to attract those who have not yet joined the Midjourney community on Discord. It is also not said how many visuals it is possible to create with free access.
Make it easier to create and attract more subscribers
For Midjourney, this is an opportunity to promote not only its young website — the company had announced this project at the end of 2023, before launching a first prototype in January — and to try to convert some of the curious into real subscribers, if they are convinced by the experience of generative AI. The first package starts at $10 per month.
The competitive landscape in which Midjourney operates is probably also not unrelated to this relaxation. Since the beginning of August, OpenAI has allowed the free creation of images for Internet users (with a limit of two images per day). OpenAI, which is behind ChatGPT, also has an image generative AI, with DALL-E.
Since March 2023 and the release of Midjourney v5, the AI system has benefited from several successive updates: v5.1 in May, then v5.2 in June. A major leap then took place in December, with the release of v6. More recently, the tool moved to version 6.1 in July. Each time, the renderings are more and more remarkable.
[ad_2]
Source link -100