Benchmark · v1, October 2026

Which Claude models tell you a product has changed?

When you ask Claude by name about a coding tool that shut down, was renamed, was bought or changed its free plan, does the answer say so?

Only tests we ran ourselves, never what makers claim.

runs counted, one answer per run
105
models ranked
3

How did each Claude model do on direct questions with search?

  1. Claude Opus 5.535 of 35
  2. Claude Fable 5.134 of 35
  3. Claude Sonnet 5.515 of 35

Answers that stated the verified change, direct questions with web search, .

Source: CookiesForAI, v1, October 2026

Higher is better

Reuse this chart

Free to reuse with credit (CC BY 4.0). Download SVG

Embed code
<figure><img src="https://cookiesforai.app/charts/benchmark--claude-models-state-a-product-change.svg" alt="How did each Claude model do on direct questions with search: Claude Opus 5.5 35 of 35, Claude Fable 5.1 34 of 35, Claude Sonnet 5.5 15 of 35."><figcaption>Source: CookiesForAI, How did each Claude model do on direct questions with search?, https://cookiesforai.app/benchmarks/claude-models-state-a-product-change</figcaption></figure>

The ranking

Which Claude models tell you a product has changed?: the ranking, higher is better
RankNameScoreDateStudy
1Claude Opus 5.5Model as shown: claude-opus-5-5100%35 of 35Read the study
2Claude Fable 5.1Model as shown: claude-fable-5-197%34 of 35Read the study
3Claude Sonnet 5.5Model as shown: claude-sonnet-5-543%15 of 35Read the study

What changed

  • : First edition, from the runs of .

All editions

Data and how to cite

Download the data: CSV (every chart of the study, this one included). Free to reuse with credit (CC BY 4.0).

How to cite
CookiesForAI, "Which Claude models tell you a product has changed?", benchmark, v1, October 2026, https://cookiesforai.app/benchmarks/claude-models-state-a-product-change/2026-10-04

Questions people ask

How do we choose what goes on a scoreboard?

Only tests we ran ourselves and published as a study, with the questions, the dates, the number of runs and the data. A board takes its numbers straight from the study's chart, so the two can never differ. We do not rank directories or other websites.

Why can our numbers differ from what makers say?

Makers usually test their own setup, with their own questions. We type real questions into the chat apps in the browser, or into Claude Code with web search on, as anyone could, on a stated day, and count what we saw. A different day, account, model or wording can give a different count, which is why every board names where it ran and the day of the runs.

What does "N of M" mean, and why do small samples matter?

M is how many runs we made and N is how many of them passed the test; each run gives one answer. The rank comes from N divided by M, and rows with the same share get the same rank. With a few dozen runs per row, a gap of one or two runs can be chance, so read close scores as close, not as a clear win.

Do we have an interest of our own?

CookiesForAI lists agent skills, MCP servers and apps, so our own pages can appear in these answers; when they do, we count them like any other site.

How do I send a correction?

Write to us with the board, the row and what you think is wrong. We check it against the saved runs and fix the board. Contact us.