the AI tool desk
New AI tools launch every day and most of the coverage is a rewritten press release. AI Tool Gazette is the opposite: hands-on reviews, honest head-to-heads, model and pricing news, and build guides written by people who ship with these tools daily and care what breaks in production.
start here
The ones that actually save time on real codebases — ranked by how they hold up past the demo.
Which frontier assistant is worth paying for in 2026, by the work you actually do.
The retrieval, chunking, and grounding choices that decide whether RAG is trustworthy or a demo.
What you'll really pay per million tokens across the major providers — and where the hidden costs hide.
tool reviews
Hands-on looks at coding assistants, agent frameworks, writing tools, and the rest — what they're good at and where they fall over.
head-to-heads
Model-vs-model and tool-vs-tool comparisons scored on the tasks people actually run, not benchmark theatre.
news & pricing
Model launches, API price changes, and policy shifts — with a plain-English take on what it means for what you build.
build guides
Practical walkthroughs: RAG, agents, evals, prompt patterns, and shipping LLM features that survive real users.
why trust this
AI Tool Gazette is an independent desk run by a Singapore team that builds and operates AI-driven software day to day. We test tools on real work before we write about them, we say when something isn't worth it, and where a link is sponsored or points to something we run, the disclosure spells it out. Want to contribute a guest article or place a sponsored post? See write for us.