Visual Generative Models
Joined April 2020
- Tweets627
- Following584
- Followers2.2K
- Likes2.3K
Ron Mokady retweeted
Ostris AI Toolkit now supports @bria_ai_ Fibo 2.0
github.com/ostris/ai-toolkit…
Being one of the top open-weight models means a lot for us
Thanks @RapidataAI for the great eval
And this is great that the data is available, will allow us to further iterate and improve 🚀
This quoted post is unavailable.
Fibo 2.0 ranks #2 amongst open-weight models in our T2I benchmark including 7M+ human judgements (700K on open-weight matchups).
1.5K prompts, all outputs, matchups, and votes can be inspected on Benchmark.ai
Ron Mokady retweeted
FIBO 2.0 by @bria_ai_ is live on each::labs on day 0 🌈
Bria’s new open-weights model for image generation and editing.
→ Generate images from text and pick your aspect ratio and output format.
→ Edit photos with text instructions or precise masks.
Try it now ⤵
Ron Mokady retweeted
Been playing with Fibo-2 Edit Turbo and it's really good at precise edits
▶️ huggingface.co/spaces/briaai…
Weights of the #1 open model in image editing just dropped on @huggingface 👀
✨coming in 2 flavors:
> Dev: huggingface.co/briaai/Fibo-2…
> Turbo: huggingface.co/briaai/Fibo-2…
Ron Mokady retweeted
Fibo 2.0 from @bria_ai_ is the #1 open weights model on AA-Image-Editing v2.1 and statistically tied with Qwen-Image-2.1 for the lead on AA-Image-T2I v2.1, a large step up from Bria's first-generation Fibo models on both leaderboards
Fibo 2.0 is Bria’s new 8B parameter open weights image model. The weights are released under a non-commercial license, which requires a separate agreement with Bria for commercial use. Bria released two sets of open weights: Fibo 2.0 Dev for Text to Image, and Fibo 2.0 Edit Turbo, a faster distilled model, for Image Editing. We benchmarked Fibo 2.0 through the Bria API at its default settings.
Fibo 2.0 is level with Qwen-Image-2.1 as the leading open weights model on both AA-Image-Editing v2.1 and AA-Image-T2I v2.1. It is a large step up from Bria's first-generation Fibo models on both leaderboards: it scores higher on every editing action and Text to Image capability we measure, with the biggest gains in Text or Symbol Edits, Text Rendering and Layout.
Fibo 2.0 costs $32 per 1k images on the Bria API. Luma UNI 1 Max ($100 per 1k images) and Qwen Image 2.0 Pro ($75) cost more and score slightly lower on both leaderboards, while Grok Imagine Image Quality scores about the same at $50. It is available through the Bria API and fal, with open weights on Hugging Face.
Congratulations to @bria_ai_ on the release!
See below for our analysis and example outputs of Fibo 2.0 on AA-Image-Editing v2.1 and AA-Image-T2I v2.1 🧵
Ron Mokady retweeted
Bria Fibo 2.0 from @bria_ai_ is live on fal.
The #1 open-weight image editing model on Artificial Analysis, with text-to-image generation and targeted editing. Trained only on licensed data, with clear commercial rights.
- Image editing with up to four input images
- Background swaps, product restaging and lighting changes
- Text-to-image generation for reasoning-heavy prompts
Wow, Fibo 2.0 ranked #1 *open weights* model in image editing 🎉
Still JSON-native control
Turbo for fast inference, Dev for fine-tuning
High-quality general editing
Better text, texture & anatomy
>>>
More details in the model cards:
huggingface.co/briaai/Fibo-2…
huggingface.co/briaai/Fibo-2…
And of course, the demos:
huggingface.co/spaces/briaai…
huggingface.co/spaces/briaai…
platform.bria.ai/image-editi…
Multi-turn editing is in the oven and coming very soon.
>>>
Ron Mokady retweeted
Multi-turn image editing is notoriously challenging: as edits accumulate, unintended changes and artifacts build up.
Not anymore!😁
Introducing ACEdit ♠️
A training approach that counteracts accumulated drift across successive image edits, preserving scene details as the changes add up.
🌐 saradorfman1.github.io/ACEdi…
1/5
Very cool research inside
Well done @omerbartal
Ron Mokady retweeted
I’ve spent the last decade working to make AI as powerful for 3D as it is for text and images - from industry, through my PhD, to leading my own research lab as a professor.
I started Thrixel to put that research directly into people's hands: a 3D creation platform for the agentic era that lets anyone build complete, interactive 3D worlds using AI.
It’s especially exciting to build Thrixel alongside people I’ve researched with for years. Together, we’re bringing fundamental research breakthroughs straight into our core engine.
Here’s a peek behind the scenes at Thrixel ↓
Thrixel is building the 3D creation engine that powers the agentic world. Born out of years of core 3D AI research, Thrixel aims to lower the barrier to entry to building in 3D and democratize the 3D creation process. Thrixel creates assets with named, editable parts, that human creators and AI agents can directly control across games, simulations, and a wide range of other applications.
See what we’re building: thrixel.com
Ron Mokady retweeted
Image analysis - everything, everywhere, all at once!
As a field, we established that more detailed captions enable better image generation models. Almost every new image model comes with some form of prompt expander, whether it is a separate model or extensive chain-of-thought. At Bria, we took this concept to the extreme. Today, we release FIBO Scene Analyzer [dev] - the model that produces the most detailed structured descriptions of images, blending free-form descriptions with precise grounding - bounding boxes, body keypoints, and segmentation masks.
What are we using it for?
Image captioning (for text-to-image training), image retrieval (fetching images by specifying objects, colors, etc.), planning editing operations, and more
Would love to see what others will do with it
[2/3]
Worried about the license?
You shouldn't be
We are granting full access to academia and independent researchers - DM me if you need assistance
Access now:
huggingface.co/briaai/fibo-s…
Last, I'm fortunate to work with @visheratin , who led this project end-to-end 👏
[3/3]
