May 10 · 8 min read · TL;DR Real LLM bills run 3 to 50 times lower than the headline per-million-token price because most input tokens come from cache. DeepSeek's deepseek-v4-flash cache read is $0.0028 per million versus $0.14 cache miss — a 50x discount. Claude Opus 4.6...
Join discussionMay 6 · 3 min read · What you'll learn How to create a dynamic pricing calculator for Claude API usage The logic behind estimating token costs across different Claude models How to build an interactive tool that updates costs in real-time Best practices for accurately f...
Join discussionMay 5 · 6 min read · I pulled up my Apify dashboard this morning before coffee. The number I was hoping for was a positive margin. The number I got was negative 33 percent. $20.94 in compute and proxy cost. $15.63 in revenue. I was paying users to run my actor. The actor...
Join discussionMay 1 · 9 min read · Recently chatted with some friends at model companies, and found that everyone is struggling with pricing. Back to the Economics Textbook In economics, price is fundamentally a supply-demand regulator. When things get expensive, demand drops and supp...
Join discussionApr 23 · 10 min read · The original issue is on GitHub #42796, and coverage is at The Register. More posts at radarlog.kr. On the afternoon of April 21, 2026, Anthropic quietly removed Claude Code from the Pro plan. The "includes Claude Code" line disappeared from the pri...
Join discussionApr 21 · 5 min read · GitHub just made Copilot more expensive — for some people. For others, they just locked the door. On April 20, 2026, Microsoft paused new signups for Copilot Pro, Pro+, and Student plans. Then they removed Opus models from the Pro tier entirely and r...
Join discussionApr 20 · 4 min read · Claude's Opus 4.6 and Sonnet 4.6 now ship with 1M token context windows at standard pricing. No premium tier. No waitlist. Just 1M tokens, available to everyone. This single change killed the moat that every RAG startup built in 2023-24. The competit...
Join discussion
Apr 18 · 8 min read · A developer running 10 million API calls monthly saw their Anthropic bill jump from $8,200 to $11,480 between February and March. The code hadn't changed. The request volume hadn't changed. The model—Claude Opus 4.7—had simply started counting differ...
Join discussion
