Research
We ask AI assistants the questions people ask, many times, and publish what they answer, which sources they use and how often the answer changes. Every study names the engines, the model versions where shown, the dates and the number of runs.
Scoreboards: Benchmarks
- Agent skill and AI tool directories: a dated comparison
A neutral, dated look at agent skill and AI tool directories, using figures read on and our Claude tests from .
- How ChatGPT, Gemini, Perplexity, Google and Grok answer software and AI tool questions
We asked ChatGPT, Gemini, Perplexity, Google AI Mode, Google AI Overviews and Grok the same 13 questions about software and AI tools, 110 times on , and recorded every name and source they gave.
- Do AI assistants know when a product changed?
In the Claude pilot, web-search runs stated the verified change in 84 of 105 direct-question labels and 36 of 84 category-question labels (21 neutral mentions excluded); a label pairs one answer with one product.
- How Claude finds agent skills it does not know, in Claude Code
In 93 runs on Claude Opus 5.5, Sonnet 5.5 and Fable 5.1 on , all 18 search runs on little-known agent skills searched the web, and 13 of 18 answers linked a skill directory page.
- AI assistants: model, search and crawler changes, dated
The engine changes tracker lists dated changes to the models, web search and crawlers of Claude, ChatGPT, Google (Gemini, Search and AI features), Grok, Perplexity, Copilot and open models, one table per assistant, each row linked to the maker's own page.