<aside> 💡
Reload this page occasionally, as it updates in real time.
</aside>
✨ Hi. This is clarinet with Midjourney. I do promptcraft community support and education. This guide was created by experienced members of the Midjourney community and is shared with Midjourney’s knowledge and support. You can find Midjourney product documentation here and the companion Midjourney #prompt-faqs to which this document belongs.
⭐ Click ▶️ to open sections. ⭐
Your v7.0 alpha mantra:
Want to see it? Say it unambiguously.
With v7.0’s improved prompt adherence, the model is more responsive than ever. If something’s missing, it might just be waiting for you to spell it out.
🎨 Want specific coaching? Try this trained GPT offered by the Midjourney #prompt-craft community. https://chatgpt.com/g/g-6781443b83888191813fbc93afa4dc64-clarinet-s-prompt-helper
For video-lovers, here’s a completely unrehearsed demo. Use the guide as a deeper dive.
Desktop? There’s a Table of Contents floating on the right hand side of this document. >>
--q) and Modes of Speed (fast, relax, turbo)--q)?“Quality” has to do with the time the model has to spend on rendering your image. If you’re into metaphors: it’s the time the image has to bake in the oven. It influences the number of steps granted to the job. The higher the value of --q, the more steps, the more time spent on the job, the more attention the model has available to give to the prompt and parameters, so the better chance of robust coherence, adherence, and aesthetics.