45 working categories — from the foundation models that anchor the field to the niche slivers we started tracking last quarter. Hand-curated taxonomy, revised every edition.
The base layer of the field — frontier LLMs, multimodal models, and the inference platforms that serve them. Where listings move slowly and stakes are highest.
The broadest engineering shelf — testing, observability, review, and the platforms around AI-assisted development. Coding assistants now rate their own category.
Prompt-to-app builders — vibe coding went mainstream this year, and the gap between prototype and shipped product is collapsing fast.
Mature category; leadership has stabilised around three frontier models, and tooling has split between consumer apps and pro pipelines.
General-purpose AI agents and the platforms that run them. Browser and computer-use agents graduated to their own category this edition.
Settled categories with established leaders. Slow-moving listings, high-stakes choices.
Fastest-growing categories — five-or-more net new entrants this edition.
Categories we started tracking in the last six editions. Coverage is still growing.
Listings shrinking — consolidation, sunset, or absorbed into broader categories.
Sponsorship is editorial-reviewed and disclosed. We won’t cover a tool we wouldn’t list in print.
Discontinued, acquired-and-sunset, acquired-and-shipping. Three outcomes, one registry.
Receive weekly updates so you can stay up-to-date with the world of AI
Receive weekly updates so you can stay up-to-date with the world of AI
The AI tools directory for discovering, exploring, and comparing the most innovative AI tools in the industry
