The blog.
Deep dives, comparisons, and roundups on the AI tools worth paying for. Only what we'd actually buy — no SEO sludge, no influencer demos.
Antigravity pricing: one number on the page, and four adjectives
Antigravity pricing starts at $0, the only number on its page. The real ladder runs $4.99 to $199.99, and Google will not say what any of it buys.
Claude Code tips: what eight months of daily use taught me
Claude Code tips from running a production pipeline on it daily for eight months, plus a controlled build test. The habits that changed my output.
Best AI for coding: 7 tools, one identical brief, measured
The best AI for coding, measured rather than listed. I gave seven tools one identical brief and used every app. All seven passed. Here is what separated them.
Lovable alternatives: 7 tools, and I gave 5 the same brief
The best Lovable alternatives, ranked by a build test rather than a feature list. Five of the seven took an identical brief. Here is what each cost.
Antigravity vs Cursor: I gave both the same brief and paid nothing
Antigravity vs Cursor on one identical build. Cursor was the fastest of seven tools, and its free tier did the whole job. Both cost me nothing.
Lovable vs Replit: same prompt, same six requirements, two bills
Lovable vs Replit on one identical brief. Both hit 6 of 6 — Lovable in 109 seconds, Replit in four minutes and 31% of its free plan. The cost model decides it.
Antigravity vs Claude Code: I gave both one brief, neither built it
Antigravity vs Claude Code, measured on one identical brief. Both refused to build unattended — the opposite of what page one says. Costs, code and verdict.
Cursor AI pricing: what the plans cost and what I paid
Cursor AI pricing runs free to $200 a month, plus metered usage. I paid $510.70 across eleven months on a plan advertised at $20. Every tier, verified live.
Google Antigravity review: it wrote its own tests
Antigravity built my six-requirement app on the free tier and met all six. Then it wrote and ran its own unit tests, which no other builder I have tested does.
OpenAI Codex review: it deleted the rule that made it wrong
OpenAI Codex renamed 33 references across 7 files and 3 languages in a single run. Then it edited my rules file, ungated, deleting the line that made it wrong.
Replit alternatives: 8 picks, and what a real build cost
Eight Replit alternatives ranked by job. I ran one identical prompt on three: Bolt built it for $0.25, Lovable $0.30, Replit would not say what it cost.
Replit pricing: what $20 of credits actually buys
Replit cut Core to $20 in February and most guides still say $25. A credit is exactly one dollar, and here is what a real measured build cost.