NewsClaude Sonnet 5 vs Opus 4.8: When to Pay MoreSonnet 5 hits ~91% of Opus 4.8's agentic-coding score at 40% of the price. A routing rule for agents: default to Sonnet 5, and when Opus still earns it.Jun 30, 2026·6 min
NewsClaude Science: Why the Model Isn't the MoatAnthropic's Claude Science runs the same Opus 4.8 you already have. The product is the workflow layer, not the model. Here's why that matters for builders.Jun 30, 2026·7 min
AI & MLVercel Eve: An AI Agent Is Just a Folder of FilesVercel's open-source Eve makes an AI agent a folder of files: instructions, tools, and skills that ship as a durable service. How it works and when to use it.Jun 24, 2026·9 min
AI & MLBuild a Simple AI Agent in PythonBuild a simple AI agent in Python: the think-act-observe loop, tool use, and memory, so the model can take steps, not just answer once.Jun 21, 2026·10 min
AI & MLTool Calling: Let the LLM Use Your FunctionsLet an LLM call your Python functions with tool calling: define tools, run the call loop, and feed results back so the model can finish the job.Jun 20, 2026·8 min