World API/EVENT SEARCH

For a day

Find semantically related events inside one dated World archive slice.

POST/api/search/{date}/semantic

Authorizations

Bearer
Authorizationstringrequiredheader

Required Bearer API key. Programmatic access to the World API needs a key; browsing nosible.world in a browser does not.

Example: Bearer nos_sk_...

Parameters

query & path
datedaterequiredpath

Archive date in YYYY-MM-DD format.

Example: 2026-07-20

Body

application/json
querystringrequiredbody

Natural-language semantic query.

Example: AI chip supply constraints

limitintegerbody

Maximum events to return.

Example: 20

embedding_modelenumbody

Embedding model. The supported value is openai.

Example: openai

Related in NOSIBLE World

World · Map ↗Open this result set on the event map, filtered to the same windowWorld · Trend ↗Plot the same query as a historical event signalWorld · Chaos ↗Watch the unfiltered crawl the ontology is built from

How to use this endpoint

guidance

Use semantic search for concept-level recall when exact keywords are too brittle. Keep the query focused and compare similarity scores across a consistent date and limit. The route returns the same lite event records used by day search, plus explicit HNSW score metadata.

Response body

application/json

These are the fields you can build against. Nested names use dot notation; optional sections are called out in their descriptions.

totalinteger

Number of semantic matches returned after the similarity floor for this request.

Example: 20

countinteger

Number of event records in this response.

Example: 20

limitinteger

Applied result cap.

Example: 20

offsetinteger

Applied offset; this route currently starts at zero.

Example: 0

eventsobject[]

Lite event records ranked by semantic similarity.

Example: [{ event_id, event, signals, coverage, _score, _hnsw_distance, _cosine_similarity }]

events[]._scorenumber

Rank-derived score, calculated as 1 / (rank + 1).

Example: 1

events[]._hnsw_distancenumber

Distance returned by the HNSW index. Lower is better.

Example: 0.13

events[]._cosine_similaritynumber

Cosine similarity derived from HNSW distance. Higher is better; the endpoint applies its configured floor.

Example: 0.87

facetsobject

Facet counts; currently an empty object for this route.

Example: {}

query_took_msinteger

Total request duration in milliseconds.

Example: 84

embedding_msinteger

Time spent creating the query embedding.

Example: 21

search_msinteger

Time spent searching the date HNSW index.

Example: 48

embedding_modelstring

Embedding model requested.

Example: openai

models_usedstring[]

Embedding models used to answer the request.

Example: ["openai"]

per_index_kinteger

Candidate count requested from the per-date index.

Example: 60

min_cosine_similaritynumber

Configured similarity floor applied before results are returned.

Example: 0.35

matched_tokensstring[]

Query token list echoed by the handler.

Example: ["AI infrastructure demand"]

errorsobject[]

Non-fatal errors reported by the semantic stage.

Example: []

Responses and errors

HTTP
200Request succeeded. The response body is shown in the panel on the right.
400invalid_request — The date, filter, cursor, identifier, or request combination is invalid.
403access_denied — The requested archive date is outside the caller's World tier.
404not_found — The requested date, entity, ticker, or event does not exist in the accessible index.
410cursor_expired — The cursor was minted against a replaced index build. Restart pagination from the first page.
429rate_limited — The request exceeded the endpoint or account rate limit. Respect Retry-After when supplied.
501backend_not_configured — The World backend is not configured for this deployment.
502backend_error — The BFF could not obtain a valid response from the World backend.
504backend_timeout — The World backend or an upstream data source timed out.