There’s also workflows around non-frontier models that make sense, especially at-scale (think internationalization of a marketing campaign to 200 countries across different media formats). A lot of the same rules that apply to LLMs apply to diffusion. The same argument could be made around Codex/Astra/Sol and many people are happily paying for tools like Devin/Cognition.

But I agree with you that if you are not careful, the subsidies from big labs can eat your business.

EDIT: I hope you’re enjoying Krea 2! What GPU are you running it on?

We know users are used to using one model for everything though and have already turned our attention to the text to image side; that message is to let people know that the text to image performance is not final, we expect the final tuning of Krea 3 to beat Krea 2’s high bar there soon.

But we’re opening the beta now (after polishing the initial jagged edges in a private beta rollout), to find even more edges now, but at least now the frontier-level capabilities and new capabilities should be available for people to get a great experience.

Do you have a workspace in Krea? I’m happy to put some API credits for you to test Krea 3.

  "subjects": [
    {
      "description": "A transparent header spans the top of the teal hero. The white ANNUI wordmark sits at the upper left, followed by Product, For Agents, Resources, and Company links. For Agents is underlined as the active page. Sign in appears near the upper right beside a large white rounded Get started button.",
      "position": "top edge, full width, approximately 8% of viewport height",
      "subject": "Navigation Header"
    },
    {
      "description": "Left-aligned hero content includes the spaced eyebrow 'FOR AGENTS', a large cream serif headline reading 'Your clients’ brighter tomorrow.', supporting copy about modern financial software, real support, and helping independent agents grow their annuity business, plus a white pill-shaped 'Build with Annui' button and a circular play control labeled 'Watch for Agents (1:30)'.",
      "position": "left half, upper-middle through lower-middle, approximately 46% width and 58% viewport height",
      "subject": "Hero Copy and Actions"
    },
    {
      "description": "A monumental classical female statue dominates the center-right of the hero. The figure has a pale blue-green face, dark teal hair, a radiant gold helmet and laurel details, and layered gold and blue garments. Architectural columns and cloudy sky fill the background, creating a sculptural financial-services editorial scene.",
      "position": "center-right, extending from the top edge to the bottom of the hero, approximately 55% width and 85% viewport height",
      "subject": "Classical Statue Hero Artwork"
    },
    {
      "description": "A narrow text panel overlays the far-right architectural columns with vertically spaced uppercase text reading 'INDEPENDENT ADVICE A BRIGHTER TOMORROW', a short horizontal divider, and 'BUILT TOGETHER'.",
      "position": "right edge of hero, upper-middle through lower-middle, approximately 15% width and 42% viewport height",
      "subject": "Vertical Message Panel"
    },
    {
      "description": "A full-width warm off-white strip beneath the hero contains four evenly spaced benefit columns separated by thin vertical rules. Each column has a pale beige circular icon and dark text: 'Streamline Sales' with quote and compare copy, 'Serve Clients Better' with carrier and education copy, 'Grow Your Practice' with tools and marketing support copy, and 'Stay Independent' with client, advice, technology, and support copy.",
      "position": "below the hero, full width, approximately 18% viewport height",
      "subject": "Agent Benefits Strip"
    }
  ],

Just for context, I ran design systems at Figma for 5 years / led their first prompt->design stream before quitting to make my own startup in this space.

Image models are trained on visual creativity, color usage, whitespace, etc - all the ingredients you need for good design principles. Where they struggle is often with information density and text rendering, which LLMs excel at. You can use the best parts of both of them though - use an image model for the initial design and an LLM to convert it to code.

One thing that's very telling is with openai's sites feature, in its chain of thought it actually renders out an image first as a design reference, then builds towards that.

You just gotta use the best parts of all of them :)

And benchmarks do not account for it sadly

Also, sometimes you want to do “ugly” things on purpose or add imperfections like film grain and such

Not yet, but it's what I quit my role at Figma to make a bet on for my startup. In my experience diffusion is a far, far better creative designer than LLMs. LLMs are trained for consistent output, which makes sense for a billing function, but leads to some very obvious smell when doing frontend work (take claude design as an example - it's often very obvious what comes out of it). Image models have a bit more creativity baked in, and are trained on color usage, whitespace, and visual balance a lot more. I've found the image->html approach leads to much better results, basically using an image model as an architect.

Would love to be connected with the team for when you all are ready to finetune it. My email is j@diffui.ai

However feels like LLMs such as Opus 5.5 in particular are surprisingly adept at generating UI mockups right off the bat these days. And since they’re already code, they’re much easier to modify and tweak.

As a serial tech entrepreneur with a few decades of hands-on experience both in my own startups and also spending some years as a VP / product group owner inside an F500 acquirer, you may want to think more deeply on this. A hidden downside of modern A/B testing analytics is how easy they make it to min/max short-term gains while masking delayed and second-order effects.

In this instance, no doubt the analytics show this bumps ASP in the short-term but those analytics can't show you the downstream impacts. These tend to accrue over time creating an overall drag on the business that you have no way to identify. A good case in point is how Adobe successfully min/maxxed their way to a large portion of their customers either not trusting (or hating) them. The brand damage is now likely irreparable. Do what you must to make payroll but you're on a slippery slope. Be careful you don't let 'any dark pattern is okay, as long as it's legal and works' become status quo. The consequences can accrue faster than you can exit.

Also, Krea 1 was co-developed with BFL !

K3 open weights is being discussed internally as I type this!

it also has enterprise level considerations that aren't present in direct API access of Google’s models, rendering almost all of the go to API documentation irrelevant, so it just comes from experience to know to use model garden’s implementation