The latest development from OpenAI has brought forth a new model, aptly named o1. According to the company, this model has been designed to take its time thinking before responding, thus enabling…
Category: AI
ColPali and Byaldi for reading PDFs with images, Reflection-70B
Multi-modal documents were always a problem, but what I can read now AI developers already made a huge progress, beating traditional PDF parsers. Now we have at least 3 solutions available as…
AI programming tools
Claude, Cursor or Replit agents? It seems like the market of application creation AI tools is growing fast. After a very good Claude Sonnet application creation chat, developers moved to Cursor and…
WIP: Hugging Face cheatsheet part 1
As prerequisite we need to have Python, pip, venv installed, for Git there is a need for lfs, because Hugging Face supports files larger than 5GB in repos. To work with with…
ComfyUI v0.2.0 released
Major updates in v0.2.0:
Flux controlnet support for Xlab, InstantX controlnets
improvements to the queue UI. You can view all outputs generated by the same task now
Bookmarking in the node library + recursive expanding / collapsing node library folders
A Content Creation Assistant
The time has come for a little presentation. After regular work, I tried to merge some tools which can help me to work on this blog. I took a month to have…
Comprehensive Resource Lists for Flux Artists
The list of artists’ names for use with Flux has been compiled by a Reddit user. To access this comprehensive resource, you can visit this cheatsheet website. In addition to the direct…
Flux.1 and LORAs
Recently, Flux LORAs have begun to appear on Civitai.com, and I decided to test a few of them to see what kind of results could be achieved effortlessly using DrawThings. To my…
Ollama Tools Support merged to langchain4j
Just quick news: finally great community work in main branch , so next release will be something we wait for a long time (joking just from 3.1). For now we can probably…
Draw Things Web API
Today, I will take a deeper look at the Draw Things Web API. I couldn’t find documentation for it, but some people have already shared messages, so I haven’t had to check…