@sofialomart

Product Marketing at @magnific | Behind @upscaleconf & @AstroCulturaSEA

Málaga, ES
Joined March 2015
Let's give @magnific some love shall we?
Who are the most tasteful companies in tech? Vote now: cosign.co/questions/highest-…
4
2
21
610
I cannot live withouth Brand kit ngl
Replying to @magnific
It all started in one conversation with Claude We asked for a food brand, landed on noodles and named it LOOSE ENDS Logo, palette and illustrations came straight from Magnific One through the MCP
2
1
14
587
Sofía López retweeted
It all started in one conversation with Claude We asked for a food brand, landed on noodles and named it LOOSE ENDS Logo, palette and illustrations came straight from Magnific One through the MCP
3
3
1
9
1,431
Sofía López retweeted
We built a full brand with Magnific One + Claude Opus 5.5 using MCP Then we turned it into an ad Here are the results 🧵
13
14
3
113
8,038
Sofía López retweeted
Magnific One launched last week and you’ve already been creating nonstop Here are 8 of the posts shared on X that have impressed us most so far Don’t get left behind, share your creation in the thread 👇
7
11
3
57
5,524
Sofía López retweeted
I'm quite amazed at what @magnific new model Magnific 1 can do with graphic design. Pretty amazing in my opinion 🔥
🤖 Made with AI
13
15
1
248
11,600
Sofía López retweeted
Singapore, Madrid, LA. Three events, one creative community We brought Magnific One. You brought the questions, ideas, and energy that make this worth building Thank you to everyone who joined us and made it happen. We can’t wait to see what you create
7
9
1
38
3,590
Sofía López retweeted
Take the same direction into a new setting Keep your Brand Kit selected and change the scene in your prompt Carry the palette, styling and mood through the series Prompts in ALT
1
2
8
505
Sofía López retweeted
Your style. Your references. Your creative direction Learn how to use Magnific One to explore ideas in Draft, combine style references, and take your favorite direction to a final image Watch the 4-minute walkthrough
10
15
2
107
434,779
Sofía López retweeted
[GUIDE] How to prompt on Magnific One Real examples, prompts to remix, and small changes that make a big difference Explore the Magnific One prompting guide 🧵
30
37
21
148
223,609
Sofía López retweeted
Text doesn't have to sit on top of the image anymore These are made with Magnific One. Turning text into visual art All the prompts are below, take them 👇
15
19
195
9,391
Sofía López retweeted
No os hacéis una idea de lo feliz que estaría mi yo de 9 años si supiera que al final nos dedicamos a algo creativo. Puede parecer una tontería, pero poder ganarme la vida creando cosas es un lujo que no quiero dar nunca por hecho 🌸
Magnific One is here A short film painted frame by frame, where color, brushwork and mood do the talking Made by Melissa Diago, with the model built by creatives, for creatives
4
1
16
401
Sofía López retweeted
Contanimation's first frutidrama. And what a drama… 
 The second episode of CONTANIMATION TV, a @magnific Original, is coming soon.

#ContanimationTV #ContanimationFilms #MagnificOriginals
3
10
978
Sofía López retweeted
Magnific One
🤖 Made with AI
Magnific One A new way to bring art direction into image generation Unlimited in Magnific Desktop until October 15
1
2
13
778
Sofía López retweeted
Magnific One is just beautiful. Download the Magnific Desktop app, the model is completely free at the moment to try. Build a brand kit based on an image style or product and you can just keep iterating all in once place.
1
5
20
1,591
Sofía López retweeted
Magnific One launched yesterday, and the most useful thing I learned in my first hour is that it wants less from you, not more. Both of these posters came from a prompt that was only the headline, the list and a three-word scene description. The reason is the brand kit. I set mine up earlier today from my site, socials and fonts, and from then on every image came back in Cormorant Garamond and Space Grotesk on a black panel with my teal, yellow and fuchsia accents, without me naming any of them. Magnific’s own guidance is to describe the subject and the moment and let the model choose the light, lens and framing, which means the prompt stays short and the look stays consistent. I ran everything from Claude through the Magnific MCP, so I could upload six of my Midjourney references as style images, test two directions side by side and keep every file in one project. Each finished 2K poster cost 180 credits. Text rendering is strong but not perfect. Three of my four posters were spelled correctly on the first try, including a five-item list. The fourth misspelled the headline and added a logo I didn’t ask for, so I erased that patch of sky with Magnific’s retouch tool and set the type myself in my brand fonts. If you make content for a brand, set up the kit before anything else, because it turns a one-line prompt into an image that already looks like your work. What would you load into a brand kit first: fonts, colors or reference images?
🤖 Made with AI
1
3
33
2,376
Sofía López retweeted
Taste over everything. Defaults for creatives, not the crowd. Aesthetics are the real moat.
Today's AI models are a technical marvel. With enough expertise, you can steer them almost anywhere. But by default they have to serve many masters, and easy prompts give you bland results. Magnific is not build for "anyone." We serve creatives, and the default mode on these models is not what our users want. Months ago we started a project with these goals: 1) built on top of existing models. Leverage the existing AI models as the building block. Get better when models get better. 2) make the default generation by and for creatives. Easy to get great results. 3) trivial to stay on brand. A new model from scratch will be something that will become obsolete in weeks / months. Our approach enhances them for our users, making a useful contribution that is long-lasting and differentiated. After many false starts, we landed on a pre, post-processing, and routing layer on top of a few of today's state of the art models. We tested the model first with our employees, and later with external users, on two tests: best image for a given prompt, and best image overall (blinding the prompt). The first checks prompt adherence, the last checks image aesthetics. Magnific One came out first on both tests by a wide margin. You can use it exclusively on @magnific, test it yourself and tells us what you think.
3
6
316
Sofía López retweeted
The "One" in Magnific One stands for being #1
Today's AI models are a technical marvel. With enough expertise, you can steer them almost anywhere. But by default they have to serve many masters, and easy prompts give you bland results. Magnific is not build for "anyone." We serve creatives, and the default mode on these models is not what our users want. Months ago we started a project with these goals: 1) built on top of existing models. Leverage the existing AI models as the building block. Get better when models get better. 2) make the default generation by and for creatives. Easy to get great results. 3) trivial to stay on brand. A new model from scratch will be something that will become obsolete in weeks / months. Our approach enhances them for our users, making a useful contribution that is long-lasting and differentiated. After many false starts, we landed on a pre, post-processing, and routing layer on top of a few of today's state of the art models. We tested the model first with our employees, and later with external users, on two tests: best image for a given prompt, and best image overall (blinding the prompt). The first checks prompt adherence, the last checks image aesthetics. Magnific One came out first on both tests by a wide margin. You can use it exclusively on @magnific, test it yourself and tells us what you think.
3
14
818
Sofía López retweeted
Today's AI models are a technical marvel. With enough expertise, you can steer them almost anywhere. But by default they have to serve many masters, and easy prompts give you bland results. Magnific is not build for "anyone." We serve creatives, and the default mode on these models is not what our users want. Months ago we started a project with these goals: 1) built on top of existing models. Leverage the existing AI models as the building block. Get better when models get better. 2) make the default generation by and for creatives. Easy to get great results. 3) trivial to stay on brand. A new model from scratch will be something that will become obsolete in weeks / months. Our approach enhances them for our users, making a useful contribution that is long-lasting and differentiated. After many false starts, we landed on a pre, post-processing, and routing layer on top of a few of today's state of the art models. We tested the model first with our employees, and later with external users, on two tests: best image for a given prompt, and best image overall (blinding the prompt). The first checks prompt adherence, the last checks image aesthetics. Magnific One came out first on both tests by a wide margin. You can use it exclusively on @magnific, test it yourself and tells us what you think.
15
18
4
92
5,443