> ## Documentation Index > Fetch the complete documentation index at: https://doc.astreus.ai/llms.txt > Use this file to discover all available pages before exploring further. # Search news, themes and transcripts > What a query finds in news, market themes and call transcripts, by its words and by its meaning: each section merges a keyword search with a search of the stored document vectors (Cohere Embed v4), ranked together; `meaning` false says that section is words only (search syntax, a company + topic query, or the meaning search was unavailable). Call rows found by meaning are verified quotes with `matchedBy` "meaning". A typed question is reduced to the words that carry it (`terms`), and quoted phrases, OR and -exclusion are honoured as typed. A company and a topic ("nvda earnings") are searched within that company. A section with `relaxed` true holds looser rows, because no row had every word and meaning found nothing: all but one of the words, or the company's rows without the topic word. `failed` true means that section's search failed. News rows with the words in the headline lead. Transcripts look back 12 months. ## OpenAPI ````yaml https://astreus.ai/openapi.json get /api/v1/search/content openapi: 3.1.0 info: title: Astreus API version: '2026-11-11' description: >- Astreus is a financial data terminal that turns the world's unstructured financial disclosure into structured, traceable data. It reads what companies, exchanges and central banks publish — annual and interim reports, SEC filings, results announcements, investor presentations, earnings calls, 8-Ks, regulator notices and policy text — across the United States, the Gulf (Tadawul, DFM, ADX, Boursa Kuwait, QSE), Japan, Taiwan, Korea, the UK and Europe, and serves it as data a program can trust: standardized financial statements (annual, quarterly, semiannual), segments and KPIs, the tables inside the notes, filing sections as text, investor decks page by page, earnings-call transcripts and audio, span-anchored observations, management guidance scored against outcomes, typed 8-K events, an attributed intraday news tape, insider and 13F ownership, a people registry, macro series (the current vintage: the latest published value for each period) and central-bank text, benchmark, commodity and compute-price series, and a source-linked claim graph with declared validation limits. Use `fact-source` and document metadata to retain available provenance. Publication and entitlement gates apply per dataset; inspect availability and tool limitations. Coverage is published live: `GET /api/v1/landing/coverage` (no key) returns every market served with company counts, history depth, completeness and periods held per data type — ask it before answering "what do you cover?". `{id}` is a ticker, a numeric SEC CIK, or a platform key for a non-SEC issuer (XSAU000001) — the `cik` field of the company payload. Human-readable guides: https://doc.astreus.ai (coverage at /coverage, data catalog at /data-catalog). Tiers: `public` endpoints need no credentials. Everything else takes an **API key** — create one on the Developers page in the terminal (https://astreus.ai/developers) and send it as `X-API-KEY: pk_live_...` (or `Authorization: Bearer pk_live_...`). Signed-in terminal sessions may use their Cognito ID token instead. Rate limits and quotas are per plan (fixed UTC windows): Free — 30 req/min, 250 req/day; Build — 300 req/min, 100,000 req/month; Scale — custom. Build, Scale and prepaid credits are set up with sales (https://astreus.ai/contact), not self-serve. Responses carry RateLimit-*, X-Quota-* and, when credits pay, X-Billing-Source/X-Credits-Remaining-USD headers; a 429 carries `Retry-After` and a machine `code` of rate_limited or quota_exceeded (with sales_email). Server-side use: `/api/v1` sends no CORS headers, so a browser page on another origin cannot call it. Call it from a server, a script or an agent, and keep the key there. Versioning: `info.version` is the UTC date the contract last changed, and every response carries it in `x-astreus-api-version`. Within `/api/v1` changes are additive; a field or parameter that will go is marked `deprecated` first and kept for at least 90 days. Policy: https://doc.astreus.ai/versioning. Not investment advice. See /terms. contact: email: saif@astreus.ai termsOfService: https://astreus.ai/terms license: name: 'Proprietary: Astreus terms of service' url: https://astreus.ai/terms servers: - url: https://astreus.ai security: [] tags: - name: Search & Meta description: Entity resolution and platform metadata. - name: Company Data description: Profiles, structure, leadership, calendars. - name: Geo & Assets description: >- Physical assets on the map — what a company owns, operates or develops, where, since when, at what location precision, and the source behind each fact. - name: Financial Statements description: >- Standardized XBRL-derived statements — every line traceable to its filing via fact provenance. - name: SEC Filings description: >- The filing index and documents (SEC and exchange-published), parsed section text, and the tables inside the notes as gated grids and series. - name: 8-K Events description: >- Material events: 8-K items and exchange disclosures by item code, filterable by family, with extracted item text. - name: News description: >- The intraday news tape: filings, releases and licensed wires with source attribution, storylines and topics. - name: Transcripts & Calls description: Earnings-call transcripts, span-anchored observations, IR decks. - name: Guidance & Scorecards description: What management promised vs. what happened. - name: Ownership & Insiders description: 13F holders, Forms 3/4/5, the people registry. - name: Macro description: >- Economic series (the current vintage: the latest published value for each period) and central-bank text. - name: Market Data description: Daily bars for stocks, benchmarks, commodities and compute. - name: Intelligence description: Narratives and source-linked claims with declared validation status. - name: Analytics description: Charting over any catalog series. paths: /api/v1/search/content: get: tags: - Search & Meta summary: Search news, themes and transcripts description: >- What a query finds in news, market themes and call transcripts, by its words and by its meaning: each section merges a keyword search with a search of the stored document vectors (Cohere Embed v4), ranked together; `meaning` false says that section is words only (search syntax, a company + topic query, or the meaning search was unavailable). Call rows found by meaning are verified quotes with `matchedBy` "meaning". A typed question is reduced to the words that carry it (`terms`), and quoted phrases, OR and -exclusion are honoured as typed. A company and a topic ("nvda earnings") are searched within that company. A section with `relaxed` true holds looser rows, because no row had every word and meaning found nothing: all but one of the words, or the company's rows without the topic word. `failed` true means that section's search failed. News rows with the words in the headline lead. Transcripts look back 12 months. operationId: get-search-news-themes-and-transcripts parameters: - name: q in: query description: Search text. schema: type: string required: true - name: section in: query description: Only this section; all three when omitted. schema: type: string enum: - news - themes - transcripts - name: limit in: query description: 1–20 per section, default 8. schema: type: integer responses: '200': description: OK content: application/json: schema: type: object additionalProperties: true '400': description: >- A query parameter the route does not take (unknown_parameter), a value outside its type, enum or range (invalid_parameter), or a required one missing (missing_parameter); `param` names it. A request refused against this spec's parameter list is not metered. content: application/json: schema: $ref: '#/components/schemas/Error' '401': description: Missing or invalid bearer token. content: application/json: schema: $ref: '#/components/schemas/Error' '404': description: Unknown company, country, symbol or resource. content: application/json: schema: $ref: '#/components/schemas/Error' '429': description: Rate limited. Honour Retry-After. content: application/json: schema: $ref: '#/components/schemas/Error' security: - apiKey: [] - cognitoIdToken: [] components: schemas: Error: type: object required: - error - code properties: error: type: string description: Human-readable message. It may change; branch on `code`. code: type: string description: >- Stable machine code; every error carries one. unknown_parameter, invalid_parameter and missing_parameter come with `param`; unknown_parameter also with `allowed`. A route may add codes of its own (e.g. key_limit); treat an unlisted code by its HTTP status. enum: - unknown_parameter - invalid_parameter - missing_parameter - invalid_request - unauthenticated - upgrade_required - forbidden - not_found - conflict - too_large - unsupported_media_type - rate_limited - quota_exceeded - auth_unavailable - upstream_unavailable - internal_error param: type: string description: The query parameter at fault, when one is. allowed: type: array items: type: string description: 'With unknown_parameter: every query parameter the route takes.' securitySchemes: apiKey: type: apiKey in: header name: X-API-KEY description: >- Astreus API key (pk_live_...), created on the Developers page in the terminal. Also accepted as `Authorization: Bearer pk_live_...`. x-default: pk_live_YOUR_KEY cognitoIdToken: type: http scheme: bearer bearerFormat: JWT description: >- Amazon Cognito ID token (token_use=id) for the Astreus user pool — the terminal's own session credential. API consumers should use an API key instead. ```` This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.