Owned by Signal8 Inc., maker of Octo and Signal8 Pro. Not investment advice. About this site
stockairesearch / lab

AskEdgar Studio for SEC filing research

By , Signal8 Research. Published by Signal8 Inc.

Blind rounds 2026-11-10 and 2026-11-11 · 4th both times · Octo ranked #1 · plans checked Nov 9, 2026

AskEdgar Studio scored 158 of 160 in our blind test of SEC filing questions and ranked #4 of 6. Signal8's Octo ranked first with 160. Studio was the only one of the top four to give a wrong figure, and as part of a paid plan it finished behind ChatGPT's free plan (159). In the harder round 3 it finished fourth of the four tools that answered every question, with 132 of 160 and 5 wrong figures, 24 marks behind Octo: the round 3 table. Signal8, which makes Octo, runs this lab.

Where AskEdgar Studio fell short in the blind round

What it is, and what it costs

AskEdgar sells a filings research product with an AI agent called Studio. Its pricing page (askedgar.io/pricing, read Nov 9, 2026) lists Free at $0 (AskEdgar Classic with filing summary previews, a basic screener and limited Copilot), Pro at $59 a month (Studio agent, dilution risk and offering data, real-time filing alerts, "Standard Studio agent usage, 2x at launch"), Professional+ at $199 a month and Enterprise by quote. The page marks dilution and offering data as not included on Free, and says API usage is billed separately. We used Studio's agent on a paid account; we are not stating what that account pays.

What we measured in the blind round

Figure correct 39 of 40, cites a source 40, source contains it 39, as-of date right 40. 39 of 40 answers scored 4 of 4.

AskEdgar Studio, blind round 2026-11-10: answers that lost a mark
QuestionCompany and topicMarks of 4Blind grader's note
R2-39LTRN, Insider purchase (Form 4)2Names only Keyser (48,544) and misses Schalop, so the total is wrong; cites accession 023082 under the wrong CIK.

Blind round caveats: every question named its filing, so the top four tools finished within 2 marks; the graders were Claude instances, the same model family as one contestant; plans differed by tool. Details on the blind round page.

Blind round 3 (2026-11-11): harder questions

Round 3's 40 questions did not name the filing, so each tool had to find the right and latest one. AskEdgar Studio scored 132 of 160 and was ranked #4 of the 4 tools that answered every question; Signal8's Octo ranked first with 156 and no wrong figure. Figure correct 35 of 40, cites a source 31, source contains it 27, as-of date right 39; wrong figures 5 (1 stale). Across rounds 2 and 3 it linked the original SEC filing on 26 of 80 answers, against 77 of 80 for Octo.

AskEdgar Studio, blind round 3 2026-11-11: answers that lost a mark
QuestionCompany and topicMarks of 4Blind grader's note
R3-02XXII, Reverse splits25 splits with correct dates/ratios and 14,904,000-to-1; only source named is an internal tool call (list_corporate_actions), no document cited.
R3-05RNXT, Nasdaq status (8-K)2No, regained compliance confirmed Aug 20, 2026; no links, only 'a filing dated August 21, 2026' with no form type (and wrong date) plus internal compliance record, so no specific source.
R3-07JAGX, Reverse splits24 splits, 60/25/35/15 with dates, 787,500. It inverts the meaning of the factor in prose, but the figures are right. The only source named is an unspecified 'exchange-sourced corporate-actions record', with no filing or URL.
R3-10BFRG, Nasdaq status (8-K)2February 8, 2027 as of Oct 31, 2026. The source is the tool's own listing-compliance record, citing 'an August 20, 2026 filing' with no form type, accession or URL.
R3-12IVF, Reverse splits1Gives 3 splits and a 120x factor, missing the March 18, 2025 1-for-12; cites only a market-data API, not a document; states the window.
R3-22KITT, Reverse splits2600,233 via 8:1 (Apr 21) and 6:1 (Sep 25) splits, factor 48; source is only a market-data corporate-actions API, so there is no specific document.
R3-27SLXN, Reverse splits2Gave 3 splits (9:1, 15:1, 10:1), cumulative 1,350 for the stated window; source only 'exchange-sourced corporate-actions record', no specific document.
R3-29SGLY, Shares and float1[STALE] Gave $4,287,562 as of June 30, 2025 from the FY2025 10-K filed Oct 14, 2025 (trap/superseded); said it found no FY2026 10-K.
R3-30CISO, Nasdaq status (8-K)2Gave Dec 28, 2026 as of Oct 31, 2026; refers only to 'the company's August 14, 2026 filing' with no form type, accession or URL.
R3-32NXXT, Reverse splits2Gave 2 reverse splits (Sept 14 and Oct 30, 2026), factor 100, citing only an unnamed 'exchange-sourced corporate-actions record' with no specific document.
R3-35HEPA, Insider purchases (Form 4)2Reported only Stetz's 2,000,000 shares at $0.05 (plus 2M warrants), not the 16,500,000 total across five filers; cited the Stetz Form 4 URL.
R3-39PMTS, Insider purchases (Form 4)2Gave 18,604 shares (only two of five buyers) and could not find the Parallel49 Form 4, so the figure is wrong; it cites 'Form 4s filed 9/14/2026' (form + date).
R3-40INV, Insider purchases (Form 4)2Only 261,000 shares / ~$394,694 (two of four filers), and its range of $1.4602 to $1.59 is also wrong. It cites 'SEC Form 4 filings, filed August 21, 2026'.

Capability round (November 12, 2026): research jobs

The capability round asked each tool to do 20 research jobs that need real data or computation: backtests, market screens, structured lookups, event studies and multi-step research. AskEdgar Studio scored 18 of 40 and ranked #5 of the five complete runs (6 done, 6 partial, 8 not done, none invented). Claude ranked first with 34 and Signal8's Octo second with 28. Every grade.

AskEdgar Studio in blind round 3

Earlier: the pilot

In the pilot on 2026-11-09, which was not blind, AskEdgar Studio scored 63 of 64, one mark ahead of Octo at 62. Its one lost pilot mark came on Q07:

How many shares of preferred stock did Our Bond, Inc. (OBAI) report as outstanding in its most recent Form 10-Q, and as of what date? Cite the filing you used.

Key in the filing: 14,651,864 series preferred shares in stockholders' deficit, plus 821,796 convertible preferred shares in mezzanine equity (15,473,660 combined). Either the 14,651,864 figure with the mezzanine class noted, or the 15,473,660 total, is accepted.. AskEdgar answered (excerpt, verbatim): As of June 30, 2026, Our Bond, Inc. (OBAI) reported 14,651,864 shares of preferred stock issued and outstanding in its most recent Form 10-Q, filed August 14, 2026. Source: Form 10-Q filed 2026-08-14, Note 8 "Composition of stock capital." The total is the sum of these series: SERIES ISSUED AND OUTSTANDING (JUNE 30, 2026) ...

Scored 3 of 4 in the pilot. Full answer in the pilot CSV (dashes shown as hyphens here).

Where it can fail on filings

Both of its misses across the two rounds left part of the answer out: one preferred class (Q07 above) and one of two insiders (R2-39), which made the blind-round total a wrong figure. On pilot Q06 it said it could not confirm a figure against the filing text and told the reader to check the fee exhibit, which is a fair habit. In Signal8's earlier 33-question internal test, Studio's answers had four wrong figures and two omissions; that test is Signal8's own, scored by Signal8, and its Studio answers date from Nov 2 (details).

How to check an AskEdgar answer against the filing

  1. Use the sec.gov link it prints and read the sentence it quotes.
  2. When it says a figure is "unverified", do what it asks: open the cover page or fee exhibit.
  3. For class or insider totals, list every class or every Form 4 yourself and add them.

Our top pick for SEC filing facts · joint top score in both blind filing rounds

Signal8 Octo, joint top score on SEC filing facts

Why Octo for filing questions: the top score in filing rounds 2 and 3 (316 of 320, shared with Claude, first on the published tie-break), zero wrong figures in 80 questions, and Octo cited the SEC filing itself on 77 of 80 answers and never relied on a third-party site. In the capability round (20 research jobs) it ranked second with 28 of 40, behind Claude (34) and ahead of ChatGPT (26). Claude vs ChatGPT vs Octo, side by side.

Signal8 Octo and AskEdgar Studio are the two filing-specialist tools in our tests. In the blind round Octo ranked first with 160 and no wrong figure; AskEdgar Studio ranked #4 with 158 and one wrong figure.

Signal8 Octo vs AskEdgar Studio: what we measured

  • Ranking: Octo ranked first of 6 in the blind round; AskEdgar Studio ranked #4. The rule and every score.
  • Blind round: Octo 160 of 160, AskEdgar Studio 158; wrong figures 0 against 1 (R2-39).
  • Blind round 3: Octo 156 of 160, AskEdgar Studio 132; wrong figures 0 against 5.
  • Signal8's internal test: Octo won 30 of 33 questions, scored by Signal8 with the scorer knowing the tool.

Where AskEdgar Studio is better, or who should pick it: it cited a source on every blind-round answer, and it is an established filings product with a free Classic tier.

Signal8 makes Octo and runs this lab, so read its results beside the other rows. Octo and Claude scored the same in both filing-accuracy rounds; Octo is first there on Signal8's published tie-breaks, set after grading (round 2: Signal8's strict re-score of flagged calls; round 3: the two rounds' record of linking the original filing). Octo is included in Signal8 Pro at signal8.ai/octo.

See Signal8 Octo