The Form 497K1 Files Dataset is a closed historical archive of voluntary mutual-fund "fund profile" filings submitted to EDGAR under the original Rule 498(b) of the Securities Act of 1933. Each record is one EDGAR submission tagged with the [497K1](https://www.sec.gov/rules-regulations/1998/03/new-disclosure-option-open-end-management-investment-companies) form type — a stand-alone, plain-English profile delivered by a registered open-end management investment company in lieu of, or alongside, the full statutory prospectus. The dataset spans the entire operational life of the form, from June 1, 1998 (the effective date of the original Rule 498) through March 30, 2009, when SEC Release No. 33-8998 rewrote Rule 498 and replaced the voluntary profile with the mandatory summary prospectus. Because the submission type was suspended on that date, the dataset is permanently bounded and no new 497K1 records will ever be filed.
Programmatically retrieve the full list of dataset archive files, download URLs and dataset metadata.
Dataset Index JSON API
Download the entire dataset as a single archive file.
Download Entire Dataset:
Download a single container file (e.g. monthly archive) from the dataset.
Download Single Container:
Form 497K1 was introduced under the original Rule 498, adopted in 1998, which permitted mutual funds to deliver a short, plain-English fund profile to investors as an alternative to the full statutory prospectus. Rule 497(k)(1)(i) defined the EDGAR submission type used to file that profile. The profile itself is a structured summary disclosure standardized around nine substantive content blocks: investment objective and principal strategies, principal risks, year-by-year past performance, fees and expenses, fund management, share-purchase mechanics, share-redemption mechanics, distributions and tax treatment, and shareholder services and account information. Profiles are short — body sizes typically run from a few thousand bytes to a few tens of kilobytes of HTML — and self-contained: narrative-plus-tabular, free of the registration-statement scaffolding (cover page, Part B/SAI, exhibits, signatures) found in a full N-1A.
The dataset covers only registrants that affirmatively elected to deliver a Rule 498 profile; adoption was voluntary and uptake across the mutual-fund industry was modest, which is why the corpus is small relative to the broader Rule 497 universe. The dataset is distributed as ZIP archives partitioned by calendar year and month of filing date (YYYY/YYYY-MM.zip), and each accession folder inside a monthly archive contains the JSON manifest plus the primary 497K1 fund-profile HTML document.
A single record in the Form 497K1 Files Dataset is one EDGAR submission filed under the 497K1 submission type. Each record is keyed by SEC accession number and is materialized on disk as a directory whose folder name is the 18-digit, dash-stripped accession (for example, accession 0000880797-08-000003 becomes the folder 000088079708000003). Inside that directory the dataset stores exactly two artifacts: a metadata.json manifest describing the filing, and main.htm, the primary 497K1 fund-profile document.
The dataset is distributed as ZIP archives. The top-level container decomposes into per-month ZIP files partitioned first by calendar year and then by calendar month of filing date (YYYY/YYYY-MM.zip). Each monthly archive expands to a YYYY-MM/ folder, and inside that folder every accession occupies its own subdirectory. There is no quarter-level partition, no CIK-level nesting, and no cross-record index inside a month — the month folder itself is the index.
Because filings are partitioned strictly by filedAt date, an amended profile and its prior version (when both fall in the dataset window) appear in different month archives and must be linked through entities.cik, fileNo, and the embedded <R> revision markers rather than through any folder-level grouping. Because Rule 498 profile delivery was voluntary and adopted by a limited set of fund families, month folders are sparsely populated; some months contain only a handful of accessions, and downstream consumers should not assume uniform temporal coverage.
Every accession folder contains the same two files, with normalized names:
metadata.json — a flat UTF-8 JSON manifest, typically 1.5–2 KB, describing the filing's identifiers, filer entity, document inventory, and back-links to EDGAR.main.htm — the primary 497K1 document, wrapped in an EDGAR SGML <DOCUMENT> envelope, typically 15–30 KB.The original EDGAR submission may have included additional artifacts — the complete-submission .txt concatenation and any .gif graphic exhibits referenced from the HTML (logos, charts, signature images). These are enumerated inside metadata.json under documentFormatFiles as URLs back to https://www.sec.gov/Archives/..., but they are not redistributed inside the container. The file types found in the dataset are JSON manifests and SGML-wrapped HTM documents; image files are intentionally excluded, and PDF courtesy renderings appear only in rare legacy submissions.
metadata.json schemaThe manifest is a flat object whose keys mirror the SEC-API filing-index schema:
formType — always "497K1" for this dataset.accessionNo — the canonical dashed accession number (e.g., "0000880797-08-000003").filedAt — ISO-8601 timestamp with timezone offset capturing the EDGAR acceptance time (e.g., "2008-04-16T10:33:59-04:00").description — the human-readable form description, typically the literal string "Form 497K1 - Profiles for certain open-end management investment companies, [Rule 497(k)(1)(i)]".linkToFilingDetails — direct URL to the primary HTML document on EDGAR.linkToTxt — URL to the complete-submission text file on EDGAR.linkToHtml — URL to the EDGAR filing-index page (...-index.htm).linkToXbrl — empty for this dataset; 497K1 predates the risk/return-summary XBRL regime introduced for summary prospectuses in 2011, so no record carries inline XBRL or accompanying instance documents.id — a 32-character hex hash uniquely identifying the index record.documentFormatFiles — an array describing every file in the original EDGAR submission. Each entry carries sequence, size (in bytes, encoded as a string), documentUrl, type (e.g., "497K1" for the primary document, "GRAPHIC" for image exhibits, or a blank string for the complete-submission .txt), and an optional description.dataFiles — an array of structured data files; empty for every 497K1 record.entities — an array of filer entities. Each entity object contains companyName with the role appended in parentheses (e.g., "FIDELITY UNION STREET TRUST II (Filer)"), the zero-padded 10-digit cik, irsNo, fileNo (the 1933-Act file number, typically of the form 033-XXXXX), act (the registering Securities Act, almost always "33"), type (matching the form type), filmNo (the EDGAR film number), fiscalYearEnd as MMDD, and optionally stateOfIncorporation. A single 497K1 may list multiple filer entities when one profile covers several series of one trust or several related trusts.The manifest therefore supplies enough information to reconstruct the original EDGAR submission inventory even though only the primary HTML is materialized on disk; the auxiliary .gif graphics and the complete-submission .txt remain reachable only via their documentUrl back-links.
main.htm)The primary document is not a clean stand-alone HTML file. It is an EDGAR SGML <DOCUMENT> envelope containing an inner <HTML> body. The first lines of every main.htm follow the canonical EDGAR submission header:
1
<DOCUMENT>
2
<TYPE>497K1
3
<SEQUENCE>1
4
<FILENAME>main.htm
5
<TEXT>
6
<HTML>
7
<BODY BGCOLOR="#ffffff">
8
... fund profile body ...
9
</BODY>
10
</HTML>
11
</TEXT>
12
</DOCUMENT>
The <TYPE>, <SEQUENCE>, <FILENAME>, and <TEXT> lines are SGML headers carried over from the EDGAR submission format and are not part of the HTML document tree. Any HTML parser applied to the file must either strip the envelope or be configured to parse only the inner <HTML>...</HTML> region. The dataset normalizes the on-disk filename to main.htm regardless of what the filer originally named the document; the original EDGAR-supplied name is preserved in the <FILENAME> header line.
Within the inner HTML, the profile is laid out as a sequence of <H1> headings introducing prose paragraphs, bulleted lists (<UL>/<LI>), and tabular blocks (<TABLE>). The standard ordering reflects the disclosure schedule prescribed by the original Rule 498:
<TABLE> of year/return pairs since the bar-chart graphic itself is not embedded), followed by an average annual total returns table showing 1-, 5-, and 10-year (or since-inception) returns benchmarked against an index. Standardized "best quarter / worst quarter" disclosure usually accompanies the bar-chart data.Body markup is plain late-1990s/early-2000s HTML: paragraph blocks, headings, bold/italic emphasis, tables of two to a dozen rows, and bulleted lists. There are no embedded scripts or stylesheets; the <BODY BGCOLOR="#ffffff"> opener is representative of the era in which these documents were authored.
The fund-profile body may contain <R> ... </R> sequences (the HTML-entity-encoded form of <R>...</R>). These are not HTML tags but EDGAR convention markers used by filers to flag text revised in the current amendment relative to a prior filing. They appear inline in paragraphs and table cells — for example, a date stamp like <P ALIGN="CENTER"><R><B>April 16, 2008</B></R> indicates that the date string is the revised content. Because the entity encoding causes them to render as literal angle-bracketed text rather than as HTML elements, a DOM-level extractor sees them as text nodes; tooling that diffs or extracts profile text needs either to strip these markers or to use them as a change-detection signal.
Image exhibits — corporate logos, performance bar charts, signature images — are referenced inside main.htm only by descriptive captions wrapped in <FONT SIZE="-1">(fidelity_logo_graphic)</FONT>-style placeholders or by table cells positioned where the chart would render. The actual .gif files are listed in documentFormatFiles with type: "GRAPHIC" and remain accessible at their EDGAR URLs but are not materialized inside the container.
A record's materialized content is the JSON manifest plus the SGML-wrapped 497K1 HTML document. The HTML carries the entire substantive disclosure: the nine profile sections above, the date of the profile, the names of the funds covered, the registrant trust identification, the share classes, and any inline amendment markers. The manifest carries the form-level metadata: accession number, filing timestamp, registrant entity (or entities) with CIK and 1933-Act file number, the full original-submission document inventory, and back-links to the EDGAR -index.htm page and the complete-submission .txt.
Several artifacts that exist in the original EDGAR submission are intentionally not redistributed inside the per-record folder: graphic exhibits (.gif files for logos and charts), the complete-submission .txt concatenation, and any .pdf courtesy renderings. These remain enumerated in documentFormatFiles with their EDGAR URLs but must be retrieved from EDGAR directly if needed. There are no signature pages, exhibit indices, registration-statement covers, statements of additional information, financial statements, or auditor consents in a 497K1 record — those belong to the underlying N-1A registration statement, not to the stand-alone profile filed under Rule 497(k)(1)(i).
The 497K1 form had a relatively short and regulatorily stable life. The original Rule 498, adopted in March 1998 and effective June 1, 1998, defined the fund-profile content schedule that maps to the nine <H1> sections described above. Over the form's eleven-year operational window the most material content changes were not driven by amendments to Rule 498 itself but by adjacent regulation:
The form was suspended, not retained in modified form, when the SEC adopted SEC Release No. 33-8998 on January 13, 2009 (effective March 30, 2009), which rewrote Rule 498. The new rule replaced the voluntary fund profile with the mandatory summary prospectus, filed under a different submission type ([497K](https://sec-api.io/sandbox/mutual-fund-summary-prospectuses), and later companion types), with a different content schedule organized around six standardized headings. Consequently no 497K1 filing exists in EDGAR with a filedAt after March 30, 2009, and the dataset closes at that boundary. Records filed in early 2009 are otherwise structurally identical to records filed in 2008 — there is no transitional content schedule.
Form 497K1 was an HTML-era form from inception. Filings from 1998 through 2009 are uniformly delivered as SGML-wrapped HTML inside the EDGAR <DOCUMENT> envelope; there is no plain-ASCII-only era for this form, and the iXBRL/XBRL regimes never applied to it. The risk/return-summary XBRL tagging requirement later attached to summary prospectuses (Rule 405 of Regulation S-T, effective 2011) postdates the suspension of 497K1 by two years, so linkToXbrl is empty across the dataset and dataFiles is empty for every record. Across the 1998–2009 window the only practical drift in source-file presentation is incidental: earlier filings tend to use sparser HTML markup, fewer inline style attributes, and simpler <TABLE> constructs, while later filings use more elaborate table layouts, multiple share-class rows, and richer typographic markup. The structural envelope and the nine-section body schedule remain constant.
Several anatomical features matter when extracting or analyzing records:
<HTML> and after the closing </HTML> or use an SGML-aware preprocessor.<R> / </R> markers are HTML-encoded, which means a DOM-level extractor sees them as literal text rather than element boundaries. To identify revised content, regex against the encoded form or decode entities first and then parse the decoded markup.entities may list several filers and the Investment Summary, Fee Table, and Performance sections may repeat per fund or carry per-class rows. There is no machine-readable per-fund delimiter beyond heading text and table column labels.documentFormatFiles is the authoritative inventory of what the original submission contained, not what is on disk. An audit of materialized completeness should compare entries with type: "497K1" against the on-disk main.htm; entries with type: "GRAPHIC" or blank type are expected to be absent locally.entities.cik and fileNo, plus the inline <R> markers that flag what changed in the amendment.Form 497K1 was filed by registered open-end management investment companies — i.e., mutual fund registrants whose shares were registered for continuous public offering on Form N-1A. The legal filer was the registrant itself (typically a Massachusetts business trust, Delaware statutory trust, or Maryland corporation organized as a series trust), submitting on behalf of one or more of its underlying series or share classes.
Structurally outside the 497K1 population:
Within the eligible universe, only registrants that affirmatively elected to deliver a Rule 498 profile generated 497K1 filings. Adoption was voluntary and uptake was modest, which is why the corpus is small relative to the broader Rule 497 universe.
Form 497K1 sat at the intersection of two Securities Act of 1933 rules:
The 497K1 tag specifically marked the EDGAR submission of the fund profile document itself under the original Rule 498. Sibling 497K-series tags covered procedurally distinct profile materials: 497K2, 497K3A, and 497K3B denoted other categories of profile-related submissions (e.g., profile amendments and related supplements) under the same 1998 framework. None of these is interchangeable with the post-2009 497K tag, which marks the modern summary prospectus.
There was no statutory deadline or transaction event compelling a 497K1 filing. The trigger was a registrant-level election to use the Rule 498 profile format for one or more series or classes already covered by an effective N-1A registration statement. Once elected, Rule 497 required the profile to be filed on EDGAR.
In practice, 497K1 submissions occurred when:
A 497K1 filing did not register securities; registration remained on Form N-1A. The 497K1 transmitted a Section 10 prospectus into the public record under the Rule 498 framework.
The 497K1 submission type was suspended effective March 30, 2009. The trigger was SEC Release No. 33-8998, "Enhanced Disclosure and New Prospectus Delivery Option for Registered Open-End Management Investment Companies," adopted January 13, 2009, which rewrote Rule 498 to create the modern summary prospectus regime, restructured Form N-1A to require a standardized summary section, and conditioned reliance on internet posting of fuller materials. After March 30, 2009, funds delivering a short-form prospectus did so under new Rule 498 using the simplified 497K tag.
No 497K1 records exist after March 30, 2009. The dataset is therefore permanently closed and bounded between June 1998 and March 2009; future updates reflect metadata refinements, not new filings.
497K summary prospectus is governed by revised Rule 498 with different content, delivery, and internet-availability conditions. It is not a continuation of 497K1.Because Form 497K1 sits inside a tightly clustered family of mutual fund disclosure submissions, the most useful comparisons are to the modern summary prospectus that replaced it, its sibling 1998-regime sub-codes, the broader Rule 497 umbrella, the underlying N-1A registration regime and its Rule 485 amendments, advertising filings under Rules 482/34b-1, and the Investment Company Act periodic reports for the same fund entities.
The direct successor and the form most often confused with 497K1. Filed under revised Rule 498 effective March 31, 2009, Form 497K is a mandatory-format Summary Prospectus with prescribed content order, fee table, and Rule 498(e) layered online-disclosure compliance. 497K1 was the voluntary nine-item profile under the original Rule 498(b), with no online-delivery framework and no prescribed item ordering. 497K is high-volume and ongoing; 497K1 stopped accepting filings on the day 497K began.
Sibling sub-codes within the same 1998 Rule 498(b) profile regime, distinguished only by the delivery scenario: profile alone, profile bound with the statutory prospectus, or profile used in lieu of the prospectus for purchase. They share 497K1's vintage, voluntary status, content categories, and March 30, 2009 suspension date. Studies of the 1998 profile experiment normally treat all four as one cohort; 497K1 is a single delivery-mode slice of it.
The general submission type for definitive materials filed under Rule 497(b)–(e) after a registration statement is effective: full statutory prospectuses, supplements, and stickers. Vastly larger in volume, ongoing, and content-unconstrained beyond the statutory prospectus framework. 497K1 is a narrow, format-specific subset of Rule 497 deliveries; Form 497 is where the long-form prospectus that a 497K1 profile summarized would typically be found.
A Rule 497(j) certification that no prospectus update is required for the year because no purchases or redemptions occurred at a price based on current NAV. Procedural and one-line in substance; carries no fund disclosure. 497K1 carries full investor-facing content (objectives, strategies, risks, fees, performance, purchase/redemption mechanics).
The registration statement for open-end management investment companies under the Securities Act and Investment Company Act, containing Part A (prospectus), Part B (SAI), and Part C (exhibits/undertakings). It is the legal source document. 497K1 is a downstream, distilled summary of selected N-1A Part A content delivered under Rule 497, not a registration filing itself.
Post-effective amendments to N-1A under Rule 485. 485APOS (Rule 485(a)) is filed for material changes and is subject to SEC staff review with a 60-day delay; 485BPOS (Rule 485(b)) is immediately effective for non-material updates such as annual financial-data refreshes; 485BXT (Rule 485(b)(1)(iii)) extends the effective date of a previously filed 485APOS. All three amend the registration statement. 497K1 amends nothing — it delivers a definitive profile document under Rule 497 after a 485 amendment has made the underlying prospectus effective.
Form 497AD covers advertising and sales literature filed under Rule 482 (omitting-prospectus ads) or Rule 34b-1 (sales literature deemed a prospectus), constrained by anti-fraud standardized-performance rules. NT 497 is a Rule 497-related notification of inability to timely file required Rule 497 material. Neither is a statutory disclosure delivery; 497K1 is a regulated summary disclosure document that itself functioned as part of prospectus delivery under the original Rule 498(b).
Investment Company Act periodic reports, not Securities Act prospectus filings. Form N-CSR is the semi-annual certified shareholder report (Rule 30e-1/30b2-1); Form N-Q was the quarterly portfolio holdings schedule (rescinded 2019, replaced by N-PORT); Form N-PORT is the monthly portfolio holdings report (Rule 30b1-9); Form N-CEN is the annual census of fund structure and operations (Rule 30a-1). All deliver post-period operational, holdings, or financial data. 497K1 delivers forward-looking summary disclosure to prospective investors. Overlap is limited to the fact that the same fund entities file both.
Form 497K1 is uniquely defined by four constraints that no neighboring dataset shares simultaneously: (1) original 1998 Rule 498(b) authority, (2) voluntary fund-profile content format, (3) a single delivery-mode designation within the K1/K2/K3A/K3B family, and (4) a closed window ending March 30, 2009. It is not the modern summary prospectus (497K), not the full Rule 497 umbrella (497), not the registration statement (N-1A) or its amendments (485-series), not advertising (497AD/NT 497), and not periodic reporting (N-CSR/N-Q/N-PORT/N-CEN). Adjacent datasets complement it — 497 supplies the contemporaneous full prospectus, N-1A and 485-series supply the legal source, and 497K supplies the modern analog for longitudinal studies of summary disclosure — but none substitute for 497K1 when the research target is the 1998–2009 profile experiment itself.
The dataset's user base is defined by its fixed historical scope and supports a narrow set of professionals who need the regulatory history, textual content, or structural metadata of an experimental disclosure regime that no longer exists.
Investment Company Act practitioners use the corpus to reconstruct the path from the 1998 Rule 498 profile to the 2009 summary prospectus. They pull formType, filedAt, and entities.name from metadata.json to build a chronological adoption set, then read the Investment Summary, Fee Table, and Fund Management sections to see which voluntary drafting conventions later became Rule 498(b) requirements. Output: comment letters, rulemaking-history memos, and historical exhibits supporting current filings.
Disclosure specialists at fund complexes benchmark today's summary prospectuses against pre-2009 voluntary profiles. They filter on entities.cik and entities.name to find predecessor filings, then compare the Investment Summary, Performance, Fee Table, and Buying/Selling Shares sections against current Form N-1A Items 2–8. Output: style-guide updates and plain-English review checklists.
Policy economists treat the filing set as a bounded natural-experiment sample for disclosure-simplification studies. They run readability metrics on the Investment Summary, parse the Fee Table for cost-presentation conventions, and use the Performance block to measure discretionary chart and return-table choices. filedAt anchors event windows; entities.cik links to later N-1A and 497K filings to measure persistence of disclosure choices. Output: working papers and rulemaking-docket submissions.
Finance scholars link entities.cik to flow and performance databases to test whether profile-adopting funds attracted different investors than statutory-prospectus funds. Legal scholars track how the Fund Management and Distributions and Taxes sections condensed fiduciary and tax disclosure. Consumer-protection scholars analyze the Fee Table and risk language for obfuscation versus genuine simplification. Output: peer-reviewed articles and dissertations.
E-discovery specialists on cases involving advisers, distributors, or transfer agents retrieve authenticated historical disclosure copies for events between 1998 and 2009. They use accessionNo and documentFormatFiles to produce the exact profile in circulation on a given filedAt, then compare the Performance and Fee Table sections against allegations in complaints. Output: production exhibits and expert-report appendices.
Reporters and historians covering fund-disclosure reform mine filedAt and entities.name for early voluntary adopters, then quote from the Investment Summary and Fund Management sections to illustrate how the industry experimented with plain-English drafting before the 2009 mandate. Output: retrospective features and long-form histories.
The corpus is valued for being small, bounded, and stylistically homogeneous: one form type, one rule. Analysts use the Fee Table for structured expense extraction, the Principal Risks section for risk-taxonomy tagging, and the Performance block for numeric extraction templates that transfer to the larger N-1A and 497K corpora. documentFormatFiles locates the profile body inside each accession; formType and filedAt provide stable train/test splits.
Records teams at fund administrators, transfer agents, and law firms maintain a pre-packaged complete population of 497K1 submissions instead of scraping EDGAR. They reconcile internal archives against the official record using accessionNo, filedAt, entities.cik, and documentFormatFiles, and retain the TXT, HTML, and PDF artifacts for fund recordkeeping rules. Output: archive integrity audits and regulator information responses.
The following workflows exploit the dataset's bounded scope (June 1998 through March 30, 2009) and its narrow content schedule (the nine Rule 498(b) profile sections). Each assumes a researcher or practitioner working off the per-record metadata.json plus main.htm.
Pull every Investment Summary and Principal Risks block from main.htm, score them with Flesch-Kincaid and Gunning-Fog readability metrics, and bin the results by filedAt year and entities.cik. Cross-reference the same fund families' post-2009 Form 497K summary prospectuses to measure whether mandatory simplification produced measurably shorter sentences and lower grade levels than the voluntary 498(b) profile. Output: a rulemaking-docket comment letter or working paper quantifying the readability delta between the voluntary and mandatory regimes.
Filter the corpus by entities.cik and entities.fileNo to assemble every 497K1 a complex filed during the relevant period, then sort by filedAt to identify the profile in circulation on a given alleged-misstatement date. Use the inline <R>...</R> markers inside main.htm to pinpoint exactly which Fee Table rows or Performance bar-chart entries changed between successive amendments. Output: an authenticated production set with a change-log exhibit suitable for an expert report.
Use the closed record set as a labeled training and evaluation corpus for parsers that target the two-block fee structure (Shareholder fees table followed by Annual operating expenses table). Because every record is HTML inside an EDGAR <DOCUMENT> envelope with a single 497K1 document type in documentFormatFiles, splits by filedAt year produce stable, deduplicated train/test partitions. Output: a fine-tuned extraction model that transfers to the much larger Form 497K and N-1A Item 3 fee tables.
Scan main.htm for the appearance of the "New Account Information" section and PATRIOT Act customer-identification language, indexed by filedAt. Plot first-appearance dates per entities.cik to chart how quickly the voluntary-profile cohort propagated the 2001-rule disclosure into investor-facing summaries. Output: a compliance-history note or short journal article on disclosure-rule diffusion across a defined fund cohort.
Aggregate entities.cik across all records to enumerate the closed list of trusts and series that opted into the 1998 profile regime, then join that list to CRSP, Morningstar, or N-SAR fund-flow panels. Compare flows, expense ratios, and asset growth between profile adopters and a matched non-adopter sample over the 1998–2009 window. Output: an empirical paper testing whether voluntary plain-English disclosure correlated with measurable investor behavior.
Use accessionNo, filedAt, entities.cik, and the documentFormatFiles URL list as a reconciliation key against an internal recordkeeping system. Flag any 497K1 the firm should hold under Rule 31a-2 but cannot produce, and use linkToTxt and linkToFilingDetails to retrieve the missing exhibit. Output: an archive-integrity audit report and a regulator information-request response package.
For each entities.cik that filed both a 497K1 and a later 497K, align the nine voluntary profile sections to the six mandatory summary-prospectus headings and tabulate which voluntary conventions (e.g., bulleted strategies, narrative-form risks, per-class fee rows) survived the 2009 mandate. Output: a fund-complex style guide or a CLE-grade memo on disclosure continuity across the Rule 498 transition.
Dataset Index JSON API: https://api.sec-api.io/datasets/form-497k1-files.json
This endpoint returns dataset-level metadata and the list of all container files. Metadata includes the dataset name, description, last updated timestamp, earliest sample date (1998-06-01), total records, total size, covered form types (497K1), the container format (ZIP), and the file types contained in each archive (TXT, JSON, HTML, PDF). Each entry in the containers array exposes the container's download URL, key, size, record count, and last updated timestamp. Because the Form 497K1 submission type was suspended in March 2009, the container set is fixed and does not receive new monthly additions, but you can still poll this endpoint to detect any reprocessing or backfills. No API key is required to call this endpoint.
1
{
2
"datasetId": "1f13365b-9ae0-69cd-ae02-628b053a8215",
3
"datasetDownloadUrl": "https://api.sec-api.io/datasets/form-497k1-files.zip",
4
"name": "Form 497K1 Files Dataset",
5
"updatedAt": "2026-04-15T12:24:06.781Z",
6
"earliestSampleDate": "1998-06-01",
7
"totalRecords": 317,
8
"totalSize": 3681342,
9
"formTypes": ["497K1"],
10
"containerFormat": "ZIP",
11
"fileTypes": ["TXT", "JSON", "HTML", "PDF"],
12
"containers": [
13
{
14
"downloadUrl": "https://api.sec-api.io/datasets/form-497k1-files/1998/1998-06.zip",
15
"key": "1998/1998-06.zip",
16
"size": 184231,
17
"records": 12,
18
"updatedAt": "2026-04-15T12:24:06.781Z"
19
}
20
]
21
}
Download Entire Dataset: https://api.sec-api.io/datasets/form-497k1-files.zip?token=YOUR_API_KEY
This URL returns the complete dataset as a single ZIP archive containing every Form 497K1 filing from June 1998 through the form's suspension in March 2009. The full archive is small enough to retrieve in a single download. This endpoint requires an API key.
Download Single Container: https://api.sec-api.io/datasets/form-497k1-files/1998/1998-06.zip?token=YOUR_API_KEY
Each container is a monthly ZIP archive grouped by filing year and month. Use the downloadUrl values from the index API's containers array to fetch only the months you need rather than the full dataset. This endpoint requires an API key.
The dataset covers SEC EDGAR submissions tagged with the 497K1 form type — voluntary mutual-fund "fund profile" filings made by registered open-end management investment companies under Rule 497(k)(1)(i) of the Securities Act of 1933, pursuant to the original 1998 version of Rule 498(b). It does not cover the modern post-2009 summary prospectus (497K) or any other Rule 497 submission tag.
One record is a single EDGAR 497K1 submission, keyed by SEC accession number and materialized as a folder containing exactly two files: a metadata.json manifest and main.htm, the primary fund-profile document wrapped in an EDGAR SGML <DOCUMENT> envelope. A single record may cover multiple series or share classes of one trust when the underlying profile was drafted that way.
No one was required to file it. Only registered open-end management investment companies (mutual fund registrants on Form N-1A) that affirmatively elected to deliver a Rule 498 profile generated 497K1 filings. Closed-end funds, unit investment trusts, variable insurance separate accounts, business development companies, and non-1940 Act issuers were structurally outside the 497K1 population.
The dataset covers June 1, 1998 (the effective date of the original Rule 498) through March 30, 2009 (the date SEC Release No. 33-8998 rewrote Rule 498 and replaced the voluntary fund profile with the mandatory summary prospectus). The dataset is permanently closed at that boundary; no new 497K1 records will ever be filed.
The dataset is distributed as ZIP archives partitioned by calendar year and month of filing date (YYYY/YYYY-MM.zip). Each accession folder inside a monthly archive contains a JSON manifest (metadata.json) and an SGML-wrapped HTML profile document (main.htm). The dataset-level index API also reports TXT and PDF as possible file types, but image (.gif) graphics referenced from the original EDGAR submission are not redistributed inside the container.
Form 497K1 is the voluntary, nine-section fund profile filed under the original 1998 Rule 498(b), with no online-delivery framework and no prescribed item ordering. Form 497K (summary prospectus) dataset is the mandatory-format summary prospectus filed under revised Rule 498 effective March 31, 2009, with a prescribed six-heading content order, fee table, and Rule 498(e) layered online-disclosure compliance. 497K is high-volume and ongoing; 497K1 stopped accepting filings on the day 497K began.
No. Form 497K1 predates the risk/return-summary XBRL regime that was later attached to summary prospectuses under Rule 405 of Regulation S-T (effective 2011). The linkToXbrl field is empty across the dataset and dataFiles is empty for every record.