The most popular AI coding tool is one you have not heard of
It has more stars than Claude Code, more than Codex, more than Gemini CLI. It is also the one nobody advertises.
It has more stars than Claude Code, more than Codex, more than Gemini CLI. It is also the one nobody advertises.
Every major provider sells the same model twice, at two prices. The cheaper one is identical in every way except when you get the answer.
Ask a model something, then ask again. The wording will differ, and occasionally the answer will too. That is not a bug being fixed.
It is roughly one four-thousandth the size of a frontier language model, and it sounds better than systems a hundred times larger. Speech turned out to be an easier problem than anyone assumed.
Total size and running size stopped being the same number, and that changes which model you can actually run.
It has 250 million downloads, it is 22 million parameters, and it will never write you a sentence. It may also be doing more real work than anything else on the hub.
One is archived. One has vanished entirely. Both had tens of thousands of users who probably do not know yet.
The most popular image model on the internet cannot legally be used to make something you sell. Nobody mentions this.
The company that spent years arguing against open models now ships them under the most permissive licence there is.
There is no catch in the pricing column. There is a catch somewhere else, and it is the oldest one there is.
Everyone calls it the same thing: using an AI model. The price list says these are not remotely the same product.
OpenAI's newest flagship and Anthropic's newest flagship are priced identically, to the cent. That is not a coincidence, and it tells you something about where the market has gone.
The phrase was chosen in 1955 partly to avoid an argument with a rival researcher — and its inventor spent decades watching people stop calling things AI as soon as they worked.