Earnings Call Transcript API - SEC 8-K Filings Parser & Monitor

Get earnings call transcripts with per-speaker tagging and separated Q&A, plus SEC 8-K filings parsed item by item from EDGAR, as clean structured JSON. One API covers both the words executives said on the call and the material events they filed with the SEC.

Pay-per-result with no start fee and no subscription. A low-cost structured alternative to expensive market data subscriptions that return transcripts as one raw text blob and filings as bare links.

What you get

Earnings call transcripts (structured, not a text blob):

• Every speaker turn tagged with name, role, and role type (executive, analyst, operator)
• Prepared remarks separated from the Q&A session
• Q&A pairs: each analyst question matched with the management answers that follow, with analyst firm affiliations
• Call participants roster, fiscal quarter, call date, word and question counts
• Guidance sentences extracted from executive speech
• Finance-dictionary sentiment scores (positive, negative, uncertainty, net)

SEC 8-K filings (parsed, not just linked):

• Full item-code classification across the complete 8-K item map, 1.01 through 9.01: earnings (2.02), executive changes (5.02), cybersecurity incidents (1.05), bankruptcies (1.03), M&A (1.01, 2.01), and every other item
• Per-item parsed text from the primary document
• Press release exhibits (EX-99.1) parsed into headline, outlook section, and body
• Guidance sentence extraction with quantity and period matching, so generic optimism is filtered out
• EDGAR acceptance timestamps for latency-sensitive event studies
• Direct EDGAR links to the filing index, primary document, and press release

Full-text keyword search: search any word or phrase across 8-K filings from ALL US companies (for example "guidance withdrawal", "material weakness", "going concern") with date and item filters.

Monitoring mode: turn on onlyNew with a schedule and each run emits only newly filed 8-Ks and newly published transcripts. A metadataOnly cheap mode powers lightweight alerting.

What you do NOT get

• No audio and no real-time streaming of live calls. Transcripts typically appear within hours of the call.
• Transcript coverage spans roughly 1,000 to 1,500 mostly large and mid cap US companies per earnings quarter, with an archive back to about 2007. Every US filer's 8-Ks are covered, but not every micro cap has a transcript.
• Sentiment scores come from a deterministic finance-domain dictionary, not an LLM. Reproducible and fast, but not a language-model judgment.
• No 10-K or 10-Q financial statements and no XBRL fundamentals. This actor is about 8-K events and earnings calls.

Use cases

• LLM training and RAG pipelines: export clean question-answer pairs and speaker-tagged remarks, ready for fine-tuning and retrieval corpora.
• Event-driven trading research: track 8-K item codes with acceptance timestamps, e.g. earnings releases (Item 2.02) or executive departures (Item 5.02), the moment EDGAR disseminates them.
• Guidance monitoring: extract forward-looking sentences from press releases and calls for a watchlist, and diff them quarter over quarter.
• Cross-market screening: full-text search every US filer's 8-Ks for crisis language such as "going concern" or "restatement".
• Compliance and IR competitive tracking: follow what peer companies file and say, in a machine-readable format.

When to use this actor

	This actor	Typical market data APIs
Earnings transcripts	Speaker-tagged turns, Q&A pairs, roles	Often one raw text blob, or gated to a top tier (FMP transcripts are Ultimate-only at $149/mo billed annually)
8-K filings	Parsed items, press release text, guidance, sentiment	Usually metadata and links only
Filing search	Full-text across all US filers	Often a separate product (sec-api.io starts at $49/mo)
Pricing	Pay-per-result, no start fee, no subscription	$49 to $149+ per month, annual commitments
Structure for AI agents	JSON records, MCP-ready	Varies

If you need every ticker's transcript back to the 1990s or real-time call audio, a premium vendor is the right tool. For structured transcripts and parsed filings on demand, this actor does it for a fraction of a cent per record.

Pricing

One event, charged per record returned: a parsed 8-K filing or a structured transcript. Error rows are never charged. There is no start fee and no separate subscription.

Worked example: 100 records cost about one cent. A quarterly refresh of a 50-ticker watchlist (one earnings 8-K plus one transcript each) is 100 records.

Input

Field	Type	Required	Default	Notes
tickers	array	one of tickers / searchKeyword	["AAPL"]	Stock symbols or CIK numbers
dataType	enum	yes	both	filings, transcripts, or both
filingsLimit	integer	no	10	Max 8-K filings per ticker (1-200); caps total results in keyword mode
transcriptsLimit	integer	no	4	Max transcripts per ticker (1-40), newest first
searchKeyword	string	one of tickers / searchKeyword	-	Full-text search across ALL US filers; treats dataType as filings
eventCategories	array	no	-	Shortcuts like earnings, executive_changes, cybersecurity that map to item codes
itemCodes	array	no	-	Exact 8-K item codes, e.g. ["2.02"]
dateFrom / dateTo	string	no	last ~12 months	YYYY-MM-DD window
includeFullText	boolean	no	false	Adds complete filing, press release, and transcript text
metadataOnly	boolean	no	false	Skip document parsing; metadata and item codes only
onlyNew	boolean	no	false	Emit only records not returned by previous runs (monitoring)

Example input

Ticker mode, filings and transcripts together:

{
    "tickers": ["AAPL", "NVDA", "MSFT"],
    "dataType": "both",
    "filingsLimit": 5,
    "transcriptsLimit": 1
}

Keyword mode, screening every US filer:

{
    "searchKeyword": "guidance withdrawal",
    "dataType": "filings",
    "filingsLimit": 25
}

Sample output

Filing record (trimmed):

{
    "recordType": "filing",
    "ticker": "AAPL",
    "companyName": "Apple Inc.",
    "cik": "0000320193",
    "formType": "8-K",
    "accessionNumber": "0000320193-26-000011",
    "filedDate": "2026-04-30",
    "acceptanceDateTime": "2026-04-30T20:30:41.000Z",
    "eventDate": "2026-04-30",
    "itemCodes": ["2.02", "9.01"],
    "itemNames": ["Results of Operations and Financial Condition", "Financial Statements and Exhibits"],
    "title": "Apple Inc. 8-K: Results of Operations and Financial Condition (Item 2.02)",
    "items": [{"itemCode": "2.02", "itemName": "Results of Operations and Financial Condition", "text": "On April 30, 2026, Apple Inc. issued a press release regarding..."}],
    "pressRelease": {"headline": "Apple reports second quarter results", "outlookText": null, "text": "..."},
    "guidanceSentences": [],
    "sentiment": {"positive": 18.94, "negative": 16.57, "uncertainty": 6.31, "wordCount": 1267, "netScore": 0.067},
    "sentimentNet": 0.067,
    "url": "https://www.sec.gov/Archives/edgar/data/320193/000032019326000011/",
    "documentUrl": "https://www.sec.gov/Archives/edgar/data/320193/000032019326000011/aapl-20260430.htm",
    "pressReleaseUrl": "https://www.sec.gov/Archives/edgar/data/320193/000032019326000011/a8-kex991q2202603282026.htm"
}

Transcript record (trimmed):

{
    "recordType": "transcript",
    "ticker": "AAPL",
    "companyName": "Apple Inc.",
    "title": "Apple AAPL Q2 2026 Earnings Call Transcript",
    "eventDate": "2026-04-30",
    "fiscalQuarter": "Q2 2026",
    "fiscalYear": 2026,
    "participants": [{"name": "Timothy D. Cook", "role": "Chief Executive Officer", "roleType": "executive"}],
    "preparedRemarks": [{"speaker": "Timothy D. Cook", "role": "Chief Executive Officer", "roleType": "executive", "text": "..."}],
    "qaPairs": [{
        "question": {"speaker": "Erik Woodring", "affiliation": "Morgan Stanley", "text": "..."},
        "answers": [{"speaker": "Timothy D. Cook", "role": "Chief Executive Officer", "text": "..."}]
    }],
    "guidanceSentences": ["We expect June quarter revenue to grow..."],
    "sentiment": {"positive": 21.4, "negative": 6.2, "uncertainty": 8.9, "wordCount": 11842, "netScore": 0.637},
    "sentimentNet": 0.637,
    "speakerCount": 13,
    "questionCount": 16
}

How to get started

1. Open the actor on Apify: View on Apify Store
2. Enter one or more tickers (or a search keyword) and click Start.
3. Read your records from the run's dataset as JSON, CSV, or Excel.

8-K data comes from SEC EDGAR (https://www.sec.gov/cgi-bin/browse-edgar), the SEC's official filing system. No API key is required.

Run from the API

curl -X POST "https://api.apify.com/v2/acts/johnvc~earnings-call-transcript-api/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"tickers": ["NVDA"], "dataType": "both", "filingsLimit": 5, "transcriptsLimit": 1}'

🔌 Use this API from Claude (MCP)

Add this actor as a tool in Claude Code (free trial), Claude Cowork (free trial), or any other MCP client, via the hosted Apify MCP server. Use this actor-specific URL:

https://mcp.apify.com/?tools=actors,docs,johnvc/earnings-call-transcript-api

Setup walkthrough:

Apify MCP integration docs: https://docs.apify.com/platform/integrations/mcp

Your AI agent can then answer questions like "What guidance did NVDA give on its last earnings call?" or "Which companies filed 8-Ks mentioning material weakness this month?" with live structured data.

n8n integration

Available as an n8n community node, n8n-nodes-earnings-call-transcript-api. In n8n: Settings, Community Nodes, install n8n-nodes-earnings-call-transcript-api, then use it in any workflow. It ships Filing and Transcript resources, a full-text Search operation, and works as an AI Agent tool.

FAQ

Which companies have transcripts?

Roughly 1,000 to 1,500 mostly large and mid cap US companies per earnings quarter, with an archive reaching back to about 2007. 8-K filings cover every US public filer without exception. If a transcript is not available for a ticker, the run returns a clear uncharged note instead of failing.

How fresh is the data?

8-K records are available minutes after EDGAR accepts the filing; the record carries the exact acceptance timestamp. Transcripts typically publish within hours of the call for large caps, and up to a day or two for smaller names.

What do the 8-K item codes mean?

Every 8-K reports numbered items. The most watched: 2.02 results of operations (earnings), 5.02 executive departures and appointments, 1.05 material cybersecurity incidents, 1.03 bankruptcy, 2.01 completed acquisitions, 7.01 Regulation FD disclosures, 8.01 other events. The actor classifies all items from 1.01 through 9.01 and includes human-readable names in every record.

Is the sentiment score from an LLM?

No. Scores come from a deterministic finance-domain dictionary (positive, negative, and uncertainty word counts per 1,000 words, with negation handling). The same document always produces the same score, which matters for research reproducibility.

Why did I get fewer results than my limit?

The limits are maximums. A company may have fewer 8-Ks in the date window, or fewer transcripts than requested. In keyword mode, only filings matching your phrase count toward the limit.

Can I search filings without a ticker?

Yes. Set searchKeyword and leave tickers empty to run a full-text search across all US filers. Add tickers to narrow the search to specific companies.

Can I monitor filings on a schedule?

Yes. Enable onlyNew and add a schedule in Apify. Each run then emits only records that previous runs have not returned, which is ideal for alerting pipelines. Combine with metadataOnly for the cheapest possible polling.

Last Updated: 2026.07.08