ir

finance ir

The ir.* commands find investor-relations presentation links and read presentation or filing document text. Use this namespace when company materials outside structured filings are useful for research.

All examples use --output json so results are stable to parse in terminals, scripts, and automation workflows.

finance ir.presentations

Discover IR and investor day presentations from SEC 8-K Exhibit 99 filings.

What it does

finance ir.presentations discovers IR and investor day presentations from SEC 8-K Exhibit 99 filings. It returns ok, data, error, and warnings in JSON output. The result payload includes symbol, presentations, count, source, sources_used, notes, warnings.

When to use it

Use this command when you need to find investor presentations, investor-day decks, or earnings slide decks before extracting text from one candidate.

Behavior details: Scans recent 8-K filings for Exhibit 99 files with presentation or slides keywords. source=auto uses SEC first, then company IR fallback when SEC finds no candidates. source=all combines SEC and company IR candidates. Press releases and earnings releases are filtered unless a distinct deck/slides signal is present. confidence: high = strong presentation signal, medium = weaker or conflicting signal. kind: investor_day | earnings_presentation | ir_presentation | exhibit_99. Use ir.read url=URL to extract text from a candidate.

Usage

finance ir.presentations SYMBOL [limit=20 source=auto|sec|company_ir|all] [--output json]

Arguments

Argument	Required	Default	Accepted values	Description
symbol	Yes	None	String	Ticker symbol to query, such as AAPL or NVDA.
source	No	None	auto, sec, company_ir, all	Provider/source key accepted by this command.
limit	No	20	Integer	Maximum number of records returned.

Basic usage

finance ir.presentations IOT --output json

Example output

This output was generated with finance ir.presentations IOT --output json.

{
  "ok": true,
  "data": {
    "symbol": "IOT",
    "presentations": [
      {
        "title": "Investor Presentation of Q4 2026, PDF file, (opens in new window)",
        "date": "2026",
        "kind": "earnings_presentation",
        "source": "company_ir",
        "url": "https://s29.q4cdn.com/853855404/files/doc_financials/2026/q4/Q4-FY26-Investor-Presentation-3.pdf",
        "page_url": "https://investors.samsara.com/",
        "page_title": "Samsara Inc. - Investor Relations",
        "company_name": "Samsara Inc.",
        "symbol": "IOT",
        "confidence": "high",
        "why_matched": "IR page link fuzzy-matches 'investor presentation'",
        "warnings": [
          "document is hosted on an external CDN linked from the IR page"
        ]
      }
    ],
    "count": 1,
    "source": "auto",
    "sources_used": [
      "sec_edgar",
      "company_ir"
    ],
    "notes": [
      "SEC candidates are scored from Exhibit 99 description and filename keywords.",
      "Company IR fallback is conservative and only crawls company/IR domains.",
      "Press releases and earnings releases are filtered unless a distinct deck/slides signal is present.",
      "Confidence 'high' = strong presentation signal; 'medium' = weaker or conflicting signal.",
      "kind: investor_day | earnings_presentation | ir_presentation | exhibit_99"
    ],
    "warnings": []
  },
  "error": null,
  "warnings": []
}

Output fields

Field	Type	Description
ok	boolean	Whether the command completed successfully.
data	object or null	Command-specific result payload. It is null when ok is false.
error	string or null	Human-readable error message when ok is false; otherwise null.
warnings	array	Non-fatal warnings returned by the command.
data.count	integer	Number of records included in the adjacent result array.
data.notes	array	Additional notes that affect interpretation.
data.notes[]	string	Additional notes that affect interpretation.
data.presentations	array	Investor-relations presentation records.
data.presentations[]	object	Investor-relations presentation records.
data.source	string	Provider or source identifier for the returned data.
data.sources_used	array	Source families used to build the result.
data.sources_used[]	string	Source families used to build the result.
data.symbol	string	Ticker symbol used by the command.
data.warnings	array	Non-fatal warnings returned by the command.
data.presentations[].company_name	string	Company name returned by the provider.
data.presentations[].confidence	string	Match confidence for the presentation candidate.
data.presentations[].date	string	Presentation or filing date when available.
data.presentations[].kind	string	Presentation or document kind.
data.presentations[].page_title	string	Title of the linked page.
data.presentations[].page_url	string	Source page URL.
data.presentations[].source	string	Provider or source identifier for the returned data.
data.presentations[].symbol	string	Company ticker associated with the presentation.
data.presentations[].title	string	Presentation title or link text.
data.presentations[].url	string	Source URL.
data.presentations[].warnings	array	Non-fatal warnings returned by the command.
data.presentations[].warnings[]	string	Non-fatal warnings returned by the command.
data.presentations[].why_matched	string	Reason a record matched the requested query.

finance ir.read

Extract text from an IR presentation exhibit URL.

What it does

finance ir.read extracts text from an IR presentation exhibit URL. It returns ok, data, error, and warnings in JSON output. The result payload includes url, text, char_count, returned_chars, truncated, format, pages, warnings.

When to use it

Use this command when you need to extract text from a presentation URL returned by finance ir.presentations or from a filing-hosted document URL.

Behavior details: HTML exhibits/pages are fetched and parsed to plain text. PDF extraction uses pypdf and returns page-level text when possible. ocr=auto or ocr=force uses the default PaddleOCR/PP-StructureV3 stack. Pass the url from ir.presentations output.

Usage

finance ir.read url=URL [max_chars=12000 ocr=off|auto|force] [--output json]

Arguments

Argument	Required	Default	Accepted values	Description
url	Yes	None	URL	Document URL to fetch and read.
max_chars	No	12000	Integer	Maximum text characters returned.
ocr	No	None	off, auto, force	OCR mode for PDFs or image-heavy documents. Use off for text-native documents, auto to fallback when text extraction is weak, and force to run OCR directly.

Basic usage

finance ir.read url=https://www.sec.gov/Archives/edgar/data/320193/000032019326000013/aapl-20260328.htm ocr=off max_chars=800 --output json

Example output

This output was generated with finance ir.read url=https://www.sec.gov/Archives/edgar/data/320193/000032019326000013/aapl-20260328.htm ocr=off max_chars=800 --output json.

{
  "ok": true,
  "data": {
    "url": "https://www.sec.gov/Archives/edgar/data/320193/000032019326000013/aapl-20260328.htm",
    "text": "aapl-20260328\nfalse\n2026\nQ2\n0000320193\n--09-26\nP1Y\nP1Y\nP1Y\nP1Y\nhttp://fasb.org/us-gaap/2025#LongTermDebtCurrent http://fasb.org/us-gaap/2025#LongTermDebtNoncurrent\nhttp://fasb.org/us-gaap/2025#LongTermDebtCurrent http://fasb.org/us-gaap/2025#LongTermDebtNoncurrent\nP329D\nxbrli:shares\niso4217:USD\niso4217:USD\nxbrli:shares\nxbrli:pure\naapl:Customer\naapl:Vendor\n0000320193\n2025-09-28\n2026-03-28\n0000320193\nus-gaap:CommonStockMember\n2025-09-28\n2026-03-28\n0000320193\naapl:A1.625NotesDue2026Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A2.000NotesDue2027Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A1.375NotesDue2029Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A3.050NotesDue2029Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A0.500Notesdue2031Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A3.600NotesDue2042M",
    "char_count": 98063,
    "returned_chars": 800,
    "truncated": true,
    "format": "html",
    "pages": [
      {
        "page": 1,
        "text": "aapl-20260328\nfalse\n2026\nQ2\n0000320193\n--09-26\nP1Y\nP1Y\nP1Y\nP1Y\nhttp://fasb.org/us-gaap/2025#LongTermDebtCurrent http://fasb.org/us-gaap/2025#LongTermDebtNoncurrent\nhttp://fasb.org/us-gaap/2025#LongTermDebtCurrent http://fasb.org/us-gaap/2025#LongTermDebtNoncurrent\nP329D\nxbrli:shares\niso4217:USD\niso4217:USD\nxbrli:shares\nxbrli:pure\naapl:Customer\naapl:Vendor\n0000320193\n2025-09-28\n2026-03-28\n0000320193\nus-gaap:CommonStockMember\n2025-09-28\n2026-03-28\n0000320193\naapl:A1.625NotesDue2026Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A2.000NotesDue2027Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A1.375NotesDue2029Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A3.050NotesDue2029Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A0.500Notesdue2031Member\n2025-09-28\n2026-03-28\n0000320193\naapl:A3.600NotesDue2042M",
        "char_count": 98063,
        "returned_chars": 800,
        "truncated": true
      }
    ],
    "warnings": []
  },
  "error": null,
  "warnings": []
}

Output fields

Field	Type	Description
ok	boolean	Whether the command completed successfully.
data	object or null	Command-specific result payload. It is null when ok is false.
error	string or null	Human-readable error message when ok is false; otherwise null.
warnings	array	Non-fatal warnings returned by the command.
data.char_count	integer	Total extracted character count before truncation.
data.format	string	Document format used by the reader.
data.pages	array	Page metadata or page count.
data.pages[]	object	Page metadata or page count.
data.returned_chars	integer	Characters included in this response.
data.text	string	Extracted document text after truncation.
data.truncated	boolean	Whether the command truncated the result.
data.url	string	Source URL.
data.warnings	array	Non-fatal warnings returned by the command.
data.pages[].char_count	integer	Total extracted character count before truncation.
data.pages[].page	integer	Page number when available.
data.pages[].returned_chars	integer	Characters included in this response.
data.pages[].text	string	Extracted text for that page or page-like chunk.
data.pages[].truncated	boolean	Whether the command truncated the result.