Knowledge base

From question to answer

Every step is an object in the knowledge base. As of 22 Aug 2026, GPT-5.6, Plus account, browser interface.

1. Session. The client loads /backend-api/me, /settings/user, /models, /system_hints and about 50 more endpoints → backend-api-settings-user, eclipse, backend-api-system-hints. The tracker's context snapshot (platform.*) comes from here.
2. Sentinel. Before every message sentinel/chat-requirements/prepare and finalize, Turnstile and proof-of-work in the browser → backend-api-sentinel-chat-requirements-prepare. The reason the runner drives the UI rather than the API.
3. Send. POST /f/conversation with 17 payload keys including model, force_parallel_switch, paragen_cot_summary_display_override → force_parallel_switch (send payload). The reply arrives as an SSE stream → UI event types in the stream (Card, Row, Pressable …).
4. Search decision. Does the model search or not? Formerly readable as Sonic probabilities, today only from the outcome → Search decision, Search rate.
5. Fan-out. A code message to web.run, line-based pipe syntax: mode, query, time window, domain → web.run syntax (pipe format), Query fan-out, Freshness window, Domain restriction, Fan-outs per question.
6. Retrieval. In-house index or Google scrape, recognisable by excerpt length → Retrieval pipelines (Labrador, scrape), Snippet median (pipeline signature). Result: search_result_groups, 14–22 URLs grouped by domain → grouped_webpages / search_result_groups.
7. Reading. Opening individual hits (open|turnXsearchY), so far only in thinking mode → open (opened results), Opened hits.
8. Answer. Text with references: inline links → url (inline reference), entities → entity (entity markup), widgets → Widgets (genui, dil, map, nav_list), source list → sources_footnote. Numbers → Citations per answer, Entries in the sources footnote, Widget rate.
9. Fetch. The client loads the finished object: since 21 Aug 2026 GET /backend-api/conversations/{id} as a paginated messages array → /backend-api/conversations/{id}. What is missing there (fan-outs, Sonic) lives on in the stream.

What the client does not see and this map therefore does not show: providers behind web.run, reading cache, fusion/ranking, crawlers. For those, see the external evidence at Retrieval pipelines (Labrador, scrape) and segonzac-2026-08-17.

Ten points where ChatGPT decides on visibility

In the order they apply. Each is a filter; what fails the first never sees the others.

1. Search decision — Does ChatGPT search at all? 69 % of our questions, yes. Evergreen knowledge, health, long-tail problems: no. No website helps there.
2. Query fan-out — Is the brand already in the query? ChatGPT writes product names into the fan-outs before it searches. Named brands are cited 33× more often than retrieved ones.
3. Domain restriction — Is the domain on the list? For 8 of 18 searching standard questions ChatGPT searches only one to five domains. All others do not exist for that query.
4. Freshness window — Is the page young — or stable — enough? The window follows the fact: news 1 day, prices 7, products 30–90, stable facts (biography, specs) up to 3650 days. Outside the window, nothing is considered.
5. Retrieval pipelines (Labrador, scrape) — Which corpus delivers? In-house index (reads 200 characters from the H1, ignores the meta description) or Google scrape (Google's snippet). Topic decides.
6. grouped_webpages / search_result_groups — Is the page among the 14–22 hits? Grouped by domain; several pages of one domain compete with each other.
7. open (opened results) — Does the page get read? Opened pages are cited 74 % of the time, merely retrieved ones 7 %. In standard mode we see no opening.
8. sources_footnote — Is the page in the source list? Two to six entries. Only those are seen by the user.
9. url (inline reference) — Is there a link in the text? The only form that gets clicked. For local and navigational questions, partly from model knowledge without a search.
10. Widgets (genui, dil, map, nav_list) and entity (entity markup) — Does an object replace the website? Map, follow-up question, product carousel, entity. No link to be won there; 18 % of our runs.

Figures are from 22 Aug 2026 and are updated daily under Search rate and the other metrics.

Register

Endpoints 5

/backend-api/conversation/{id} (superseded) /backend-api/conversations/{id} backend-api-sentinel-chat-requirements-prepare backend-api-settings-user backend-api-system-hints

Fields 2

search_model_queries run.conversation_format (mapping | messages)

Feature flags 1

eclipse

Models 1

gpt-5-6 (GPT-5.6 Sol)

Retrieval pipelines 1

Retrieval pipelines (Labrador, scrape)

Reference types 4

entity (entity markup) grouped_webpages / search_result_groups sources_footnote url (inline reference)

Send payload 1

force_parallel_switch (send payload)

Sources 22

blyskal-2025-11-03 blyskal-2025-12-08 blyskal-2026-07-22 goodwin-2026-07-08 green-2025-08-09 konitzny-2025-11-19 konitzny-2026-06-22 littlegreenagency-2026-07-02 mcsweeney-2025-12-19 mohanadasan-2026-06-24 mohanadasan-2026-07-14 mohanadasan-2026-08-10 mohanadasan-2026-08-21 promptwatch-2026-08-10 ray-2026-08-17 rudzki-2026-05-05 schneider-2026-01-02 segonzac-2026-02-01 segonzac-2026-05-14 segonzac-2026-08-17 wells-2026-02-12 yesilyurt-rrf

Stream types 1

UI event types in the stream (Card, Row, Pressable …)

Terms 2

Query fan-out Search decision

web.run modes 9

business (location search) Domain restriction fast (standard web search) Freshness window image (image search) open (opened results) product (product search) slow (freshness-driven search) web.run syntax (pipe format)

Widgets 1

Widgets (genui, dil, map, nav_list)