POST api/stpassistant/run/v4/{Roomkey}?AssistantKey={AssistantKey}

V4 summary generation: identical flow to v3 (conversation reuse, checklist carry-forward, background Responses API + webhook completion) but on the gpt-5.5 high-reasoning model with the design-contract instruction so the model also returns the rich meeting-brief HTML body. V3 stays untouched; switch traffic by pointing the run proxy at run/v4 once verified.

Request Information

URI Parameters

NameDescriptionTypeAdditional information
Roomkey

globally unique identifier

Required

AssistantKey

globally unique identifier

None.

Body Parameters

None.

Response Information

Resource Description

HttpResponseMessage
NameDescriptionTypeAdditional information
Version

Version

None.

Content

HttpContent

None.

StatusCode

HttpStatusCode

None.

ReasonPhrase

string

None.

Headers

Collection of Object

None.

RequestMessage

HttpRequestMessage

None.

IsSuccessStatusCode

boolean

None.