Before the recap, confirm the tours actually landed.
You're good when (any two platforms)
Check both boxes:
- Deeper chat or research: You found a Thinking / Pro model control or a Research / Deep Research / DeepSearch entry (even if upgrade-only).
- Search or images: You found web search, image generation UI, or Imagine — and you wrote one sentence: "Platform A has ___; Platform B has / doesn't have ___."
Grok counts if you did step 6. Skipping Grok is fine if ChatGPT + Claude + Gemini (or any two of the three) pass.
Quick self-test
| Platform | Model / research found? | Search or images found? |
|---|---|---|
| ChatGPT | ||
| Claude | ||
| Gemini | ||
| Grok (optional) |
Common fixes
- Can't find Thinking / Research — check you're on desktop web, not a stripped mobile view; search the platform's help for the feature name.
- Everything says Upgrade — you still pass: note menu location +
not on my tier. - Wrong account — work vs personal Microsoft/Google/OpenAI login changes what's visible.
- Labels changed — write what your screen says; tell us via site feedback if the lesson is badly out of date.
- Claude "images" — browser extensions and third-party tools don't count; we're mapping official app menus.
Continue once both boxes are ticked on at least two platforms.