AI Search
@aisearchio
Sharing AI stuff ๐ 760K subs
Nature lover, PhD, investor, shitposter
344 Following 27.4K Followers
today, Krea 3 makes it to the top of image arenas.
- "Fast" pushes the pareto curve of speed and quality.
- "Ultra" ranks top-4, only surpassed by OpenAI models.
the only company valued under $1T in the top-10!
more below ๐
Show more
AI never sleeps. I just finished the video on MOSS, and now we have a new image EDIT model, Fibo 2.0.
Apparently it's the #1# open image edit model.
Testing it out today.
Show more
MOSS Image knowledge test
Full tutorial here
Spectrum node for MOSS Image. Up to 2.5x speedup with no noticeable quality loss
Kromoss, a new finetune of MOSS Image for even better quality and consistency
New model from Deepseek. 77 on DeepSWE!
๐ DeepSeek-V4.1-Flash-1017 is here!
Stronger benchmark performance, faster task completion, and shorter outputs than the previous versionโa Pareto improvement in performance and cost. Now live on our website, app, and API.
๐ ๏ธ Optimized for @DeepSeekHarness, Tencent CodeBuddy, WorkBuddy, and more. Upgrade to DeepSeek Harness v0.2.1-rc.1 or later for the best experience with the new model.
1/4
Show more
The choice of demo images from the MOSS Image team ๐คฃ
MOSS Image 8 step finetune is here. ~5x speedup compared with the base model (50 steps).
Now THIS is a good image model. MOSS Image, 5B Apache 2.0.
Already very uncensored!
Kling 4 is also out today. AI never sleeps
To clarify, I remain optimistic about Flux 3 Klein. Most of these issues could be solved with (corn) checkpoints or loras.
The previous Flux Klein with loras had been my go-to edit model for a while. Hopefully Flux 3 finetunes can exceed this.
Show more
Flux 3 Klein is more censored than Qwen 2.1, and has some occasional anatomy issues. Not impressed so far
Drama 3 by Fish Audio is now the top open TTS model.
Allows meta tags within your transcript to control emotion, direction, SFX. Pretty good at voice cloning too.
Show more
New TaoMate H3 lora for speeding up Minimax H3 in ComfyUI
Krea 3 is out! I've been testing it for the past few days and it's a pleasure to use.
It can do image editing with natural language, just like Nano Banana, but much faster & cheaper.
Show more
today we open the beta for Krea 3, a new frontier for image editing.
- end-to-end 2k edits in under 8s.
- matches Nano Banana Pro or GPT Image 2.5 quality.
- bounding box control and precise multi-turn edits.
learn and get access ๐
Show more
Flux 3 Klein is more censored than Qwen 2.1, and has some occasional anatomy issues. Not impressed so far
FLUX 3 Image Klein is here.
Open Weights with the quality of a large frontier model. Customizable and fast.
5.6B parameters and runs on a single consumer GPU.
Native 4K, bounding boxes, and pixel-perfect editing.
Show more
Exciting stuff from Tripothon SF Demo Day.
Creators showcased everything from playable games and immersive VR worlds to interactive installations and 3D-printed creations.
AI 3D generators like @Tripoai are becoming an essential part of the workflow.
Show more
Exciting stuff from Tripothon SF Demo Day.
Creators showcased everything from playable games and immersive VR worlds to interactive installations and 3D-printed creations.
AI 3D generators like @Tripoai are becoming an essential part of the workflow.
Show more
I hope Terrence Tao is having a nice day
This is arguably the biggest day (yet) in mathematics. OpenAI dropped HUNDREDS of math breakthroughs at once.
It seems that the bottleneck is no longer AI intelligence or compute, but rather, human capacity to process all of this.
There's so much catching up to do...
Show more
This is arguably the biggest day (yet) in mathematics. OpenAI dropped HUNDREDS of math breakthroughs at once.
It seems that the bottleneck is no longer AI intelligence or compute, but rather, human capacity to process all of this.
There's so much catching up to do...
Show more
