Agreeing with the comments, it is almost as good as Opus, for a small fraction of the price. This is a great sign, and also the confirmation that writing code by hand is going to be rarer and rarer this year.

As the title says, a cookbook on working with structured data, by people who created open source OCR and document processing tools

Seems almost too good to be true, but apparently saves about 80% of tokens. To be tested.

A small embeddable text-to-speech with impressive quality for its size. Works on mobile

And another AI tool to help with development… With other AI tools.

Well, here’s another AI coding tool to try out!

A tool to write agentic scripts using Markdown. It’s a great idea and looking at the examples seems to do the job.

A model with great performance on vision, including OCR and structured data extractor

Insightful notes on the use of AI in coding, in a nutshell, own the code and look for (high gradient) opportunities.

A Xero competitor with automatic reconciliation

A detailed first hand account of working with Claude Code.

Finally, some convergence. Weird that OpenCode is not mentioned, though.

A new library in Python designed specifically for working with LLM. It does seem quite refined in it’s thinking, and goes beyond a simple adapter or collection thereof.

It’s a good idea to use AI to review the security of your code, although you may get surprising results!

A really good presentation on the security vulnerabilities of today

An open source project from Mozilla that helps use multiple models seamlessly.

When SoTA LLMs are tasked to keep books tidy for a long period of time, they drift and the balance accuracy falls over time. This is a good example that we still need humans in the loop

An extensible templating language designed for writing prompts.

Traces, evals, prompt management and metrics to debug and improve your LLM application. Integrates with Langchain, OpenAI, LlamaIndex, LiteLLM, and more.

Your open source AI agent, automating engineering tasks seamlessly.