Marwan

Founding Team @ideogram_ai

[Marwan @marwangen]

@genena_zeyad @ideogram_ai who is joining from our team : )

[zeyad genena @genena_zeyad]

Attention Toronto ‼️

I’m hosting a 3-on-3 basketball tournament where startups will compete to prove who has the best squad in the city

Comment and I will send you the link to apply

[Kamryn Ohly @KamrynOhly]

Congrats to the team! one of the largest open weights jumps we’ve ever seen

[Marwan @marwangen]

@humphreysu8 @benaratame we should go

[Design Arena @Designarena]

https://t.co/4ZArl4QOjV

[FLORA © @floraai]

Ideogram 4.0 is the image model built for design, now where the rest of your work already lives.

Check out @ideogram_ai in FLORA.

[rob - comfyui @hellorob]

The best way to bring the composition from your head into an image → Ideogram V4 + drawing bounding boxes in Comfy.

The control here is quite unique. The model uses structured JSON so drawing bounding boxes to get the exact placement works very well.

The model only needs 12 steps (turbo), so iterating with different seeds + very impressive text rendering capability leads me to say this is a state of the art open source image model right now.

Using the 'Ideogram 4 Prompt Builder' node by Kijai.

[LovartAI @lovart_ai]

Ideogram 4.0 × Lovart: Fresh drop today🚀

[Arena.ai @arena]

In the Image Arena: open-weight Text-to-Image has a clear leader, with a tight race directly behind it:

- #1 Ideogram-4.0 Quality has set the pace this week with a score of 1204. @ideogram_ai
- #2 Hunyuan Image 3.0 by @TencentHunyuan with a score of 1151, just +1 pt ahead of Flux-2 Dev @bfl_ai at #3.
- #4 Qwen Image 2512 by @Alibaba_Qwen and #5 HiDream-O1 Image @HiDream_AI complete the top five, scoring 1128 and 1124.

The top six are represented by different labs, while Flux and Qwen provide the greatest depth across the Top 15.

[Marwan @marwangen]

For those interested in how we trained the best open weight image model 👀

[Ideogram @ideogram_ai]

Today we published a technical blog post about Ideogram 4.0 — our goal is to enable more innovation and creativity.

It's a 9.3B Diffusion Transformer trained from scratch, paired with a frozen 8B VLM as text encoder. The nf4 checkpoint runs on a 24GB consumer GPU.

Thread 🧵

[Ideogram @ideogram_ai]

Today we published a technical blog post about Ideogram 4.0 — our goal is to enable more innovation and creativity.

It's a 9.3B Diffusion Transformer trained from scratch, paired with a frozen 8B VLM as text encoder. The nf4 checkpoint runs on a 24GB consumer GPU.

Thread 🧵

[Mohammad Norouzi @mo_norouzi]

Fun fact about Ideogram 4.0: we haven't even scaled the model yet.

We're hiring! My team and I are at @CVPR for 3 days. DM me, swing by the booth, or come to our happy hour tonight: https://t.co/iOlm7PXYyq

[Marwan @marwangen]

@MoonGotArt @ReflctWillie If you care about getting the best quality you can think so : )

[Ideogram @ideogram_ai]

Our CEO @mo_norouzi was on @MTSlive

[Mohammad Norouzi @mo_norouzi]

Ryan Dahl is the most cracked engineer I know. He started Node.js and is the CEO of @deno_land now. He spent a year doing research with me and Jon Shlens and we had a hunch in 2017 that generative models are good for image upscaling.

[zeyad genena @genena_zeyad]

@ideogram_ai The ML team really locked in this winter hahah

[Lucas Beyer (bl16) @giffmana]

Huh ideogram going open weights was not on my bingo card! Nice

(Apache-2.0 code, NC weights)

[fal @fal]

Ideogram 4.0 the best open weight model is here!

[Marwan @marwangen]

Image editing next 🚀

[Ideogram @ideogram_ai]

Introducing Ideogram 4.0: the best open image model in the world.

Think it. Make it. Own it.

Download the weights, fine-tune on your own data, and run it on your hardware. Live on every Ideogram plan and the API today.

[Marwan @marwangen]

🇨🇦🫡

[salehdigital @saleh_digital]

toronto btw