In Jev, every field is generated on its own. Its documentation puts it plainly: "every question is evaluated in parallel and in isolation against the same state."
\nOur last post went one step further: with
\ndepends_on, a later field sees the answers of earlier ones. But every field still ran. An earlier answer could change what a field said, never whether it was asked.
\nNow it can. With
\nwhen, earlier answers decide which fields TypeLLM generates next, all in one API call.
\nAn inbox gets invoices, meeting requests and job applications. Each needs something different: an amount, a date, a role. Ask one question first, what kind of email this is, and let its answer decide which follow-up runs:
\nFor an invoice ("Amount due: $1,240.00 by October 15"), the API returns:
\nA meeting request takes the meeting branch instead and returns
\n"meeting_date": "2026-10-14"; a job application returns its
\nrole. Each follow-up is a typed field of its own: a number, a string, an enum.
\nTyped outputs always return a value in the field's type, even when the question makes no sense. Asked of the same invoice without
\nwhen, the follow-ups for a meeting date and a job role came back as the string
\n"null"and
\n"engineering": well-typed, and meaningless. With
\nwhen, those fields are skipped instead: they have no key in the result, and
\nresponse.skippedlists them.
\nwhenmaps earlier fields to a test of their answers:
\nNo
depends_onneeded. The fields a condition names become dependencies, so the field also sees their answers.
You could do this yourself: call once to classify, read the answer, then call again with the right follow-up. Doing it in one call is better in four ways.
\nNo glue code. No parsing the first answer, no
if, no second request to build. The branch lives in the schema, and the answers come back together, with
\nskippedsaying what did not apply.
whenis in TypeLLM 0.5.1 and the TypeLLM API. Run the email example in the playground or read the conditional fields docs.
","headings":[{"level":1,"text":"Generate only when it applies","id":"generate-only-when-it-applies"},{"level":2,"text":"Sort an email, then ask what that kind needs","id":"sort-an-email-then-ask-what-that-kind-needs"},{"level":2,"text":"Why not just ask everything?","id":"why-not-just-ask-everything"},{"level":2,"text":"What a condition can test","id":"what-a-condition-can-test"},{"level":2,"text":"Why one call, not your own code?","id":"why-one-call-not-your-own-code"},{"level":2,"text":"Try it","id":"try-it"}]}}