Kepler vs Daloopa

For investment teams who need numbers they can defend and the analysis built on them, Kepler is the stronger choice. Daloopa supplies source-linked data points on 6,000+ public companies for a spreadsheet or another AI tool to reason over. Kepler is the whole platform around those numbers: statements assembled from each filer’s XBRL with every cell cited, your firm’s ontology and definitions, calculations in code, agents that keep the work current and every surface your analysts work in.

We build Kepler. Facts about Daloopa come from its website, documentation and public reporting as of November 2026, linked inline.

The short answer

• Choose Kepler when the numbers feed a decision: comps on your own adjusted EBITDA, screens across your coverage, an earnings note by morning, a model that updates itself when the 10-Q lands. Choose it when your own documents and definitions belong in the answer, and when every figure has to open its source and its arithmetic.
• Daloopa may fit when the job is a fast, source-linked historicals feed into Excel models your team already maintains, and the rest of the workflow is settled.

Kepler vs Daloopa at a glance

	Kepler	Daloopa
What you get	The platform: data, your firm’s ontology, a deterministic engine, cited answers, agents and every surface	Source-linked historical data points, plus the Scout Excel agent
Coverage	14,000+ companies in 120+ countries; 26M+ SEC filings, 40M+ IR documents and transcripts	6,000+ public companies, up to 14 years of history
Where each number comes from	Pulled from the filing and computed by code; statements from XBRL with no model in the path	Captured by an automated tagger, then reviewed by analysts
What a citation opens	The exact page, line or cell, with formula and inputs for every calculation	The exact spot in the source behind each data point, per Daloopa
Derived numbers	Computed in code over cited inputs, formula visible	Your Excel formulas, or ones Scout writes
As filed vs as restated	Either, on any date; histories rebuilt so each line keeps one meaning	Original and restated values both kept and flagged
Your firm’s definitions	Built into your ontology, applied to every company	Daloopa’s standard taxonomy; your adjustments live in your spreadsheet
Text beyond the numbers	Every filing, transcript and deck read in full: risk factors, footnotes, guidance language	Data points, guidance and filing text through MCP
Your own documents	Notes, models, CRM and data rooms in your ontology	Outside its scope
Same question twice	The same number, with the same sources	Same data points; AI answers depend on the assistant
Work that runs on its own	Agents rerun models and comps when companies file, and send the changes	Add-in updates after a print; Scout updates a model on request
Where it works	App, Excel, PowerPoint, Word, Claude, ChatGPT, Slack, API, MCP, Python SDK	Excel add-in, data sheets, API, MCP, Snowflake, Databricks, S3
Rules and audit	One permission check on every read, every query and calculation logged	Not detailed publicly

What Daloopa is

Daloopa is a source-linked fundamentals provider. It covers 6,000+ public companies with up to 14 years of history, and every figure carries a hyperlink to the filing or presentation it came from. Its documentation describes two ways a value arrives: an automated tagging pipeline captures it soon after a filing or earnings release, and an analyst-reviewed capture follows later. When a company restates, Daloopa keeps the original and the restated value as two records, and a taxonomy layer maps each company’s own series names to standard metrics so peers can be compared.

The data reaches analysts several ways. Data sheets give a full history for a ticker; the Excel add-in links a model to Daloopa and refreshes it after a print; Scout, an Excel agent that became generally available on September 9, 2026, builds a model or updates one with a new quarter. An API, cloud delivery into Snowflake, Databricks and S3, and an MCP connector for Claude, ChatGPT and Microsoft 365 Copilot serve teams that build their own tools. ChatGPT for Financial Services includes select Daloopa data. Daloopa’s plans page lists a free tier with up to three data sheets and Core, Premium and Fundamentals API plans through sales.

What Kepler is

We build enterprise intelligence for investment firms: one living model of your firm, its data, definitions and relationships, that AI can reason over, prove and act on.

Your firm’s ontology. Our forward-deployed engineers, who built this layer before at Palantir, build it with your team. Each company, segment, peer set, holding and internal model resolves to one record, and each house term (adjusted EBITDA, organic growth, net debt to EBITDA) carries one definition that every answer uses. It updates itself as documents arrive and remembers what was true as filed, as restated or on any date. See the ontology.

Every document, read in full. Public data is loaded on day one: 26M+ SEC filings across 20+ form types with full text and XBRL, 40M+ IR documents and earnings call transcripts, market data and news for 14,000+ companies in 120+ countries. Your models, notes, CRM and data rooms connect read-only in the first week. Extraction models we train read every page at ingest and keep each fact with its value, scale, unit and passage; on held-out filings they read tagged facts at the right scale 99.4% of the time (Stop Paying to Forget).

A deterministic engine. Code pulls and computes every number. Statements assemble from each filer’s XBRL with no model in the path, and in our paper Reclassified to Conform, we show how to rebuild histories where 194 of 200 companies reshaped a statement beyond what XBRL tags track. Every derived figure runs over cited inputs with its formula visible, and each citation is checked against its passage before you see it.

Agents, playbooks and every surface. Agents start on schedules and on events such as a new 10-Q or a guidance change, acting exactly as the person who set them up. Your playbooks (the comps sheet, the earnings update, the initiation template) are saved once and applied every time. The work lands in the Kepler app, Excel, PowerPoint, Word, Claude, ChatGPT and Slack, and your engineers build on the same engine through our API, MCP server and Python SDK. One access check governs every read, every run is logged, and we hold SOC 2 Type II.

Not the right fit if: your team only wants a historicals feed for Excel models it already maintains, and has no plans to bring its own documents, definitions or analysis onto the same platform.

How each one answers the same question

Take a request a payments analyst might get on a Monday: “Gross margin and adjusted operating margin for our six payments names over the last 12 quarters, on our definition (GAAP operating income plus stock-based compensation and acquired-intangible amortization), calendarized to March quarter-ends. Flag any segment recasts, and keep it current as each company files.”

With Daloopa, the analyst pulls revenue, cost of revenue, operating income, stock-based compensation and amortization for each company into a template through data sheets or the add-in, each value linked to its source. Company-specific line names are mapped through Daloopa’s taxonomy or by hand, and the add-backs, margins and calendar shifts are written as Excel formulas, by the analyst or by Scout. Where a company restated, the analyst chooses between the original and restated records. After each 10-Q the add-in refreshes the inputs, and the analyst rechecks the formulas, the mappings and the calendar logic before the sheet goes to the PM.

With Kepler, the six names are already a coverage list in your ontology, and “adjusted operating margin” and your March calendarization already carry your definitions. The engine takes each input from the filing itself (operating income and cost of revenue from the statements, stock-based compensation from the cash flow statement with year-to-date figures turned into quarters, amortization from the notes), applies your definition and calendar rule in code and fills all 12 quarters. Every cell opens its source line, and every margin opens the formula with its inputs. Where a company recast a segment, the table shows the as-filed and restated figures side by side and flags the change. An agent reruns the table the morning each 10-Q lands and posts what moved, with the relevant line from the call, to the analyst’s Slack and Excel.

The difference shows up the second time. In Kepler the definition, the calendar rule and the recast handling are part of the firm’s ontology, so the next analyst who asks gets the same numbers without rebuilding the sheet.

“I can’t put a number in front of a client if I can’t show where it came from.” (Managing Director, multi-billion dollar fund)

When to choose Kepler

• The numbers go into a decision. Comps, IC memos, earnings notes and client decks where every figure has to open its source and its arithmetic.
• Your definitions differ from a standard taxonomy. Adjusted EBITDA, organic growth or segment mappings the way your firm defines them, applied to every company.
• You need the text as well as the numbers. Risk-factor changes, footnote language, guidance quotes and non-GAAP reconciliations, read in full and returned as complete lists.
• Your coverage runs past 6,000 names. Kepler covers 14,000+ companies in 120+ countries.
• Your own documents belong in the answer. Internal models, research notes, CRM history and data rooms next to the filings.
• Work should run without a click. Models and comps that update when a company files, alerts with the exact quote, digests before the morning meeting.
• Your team works in many tools. Excel, PowerPoint, Word, Claude, ChatGPT, Slack, or your own apps on our API.
• Engineering wants to build. Firms with engineers build their own agents on Kepler’s data, ontology and engine, and skip the data layer that took us 17 months and 109 releases.

When Daloopa may fit

• A historicals feed is the whole job, into Excel models your team already maintains.
• You’re building your own tools and want a fundamentals dataset by API or in Snowflake.
• Your workflow is settled and you only need the inputs refreshed after each print.

For more options, see Daloopa alternatives, Daloopa vs Canalyst, Kepler vs AlphaSense and our guide to checking AI numbers.

Put the same request to both. Start with Kepler at app.kepler.ai, or book a demo on your own coverage list and definitions.

Frequently asked questions

What is the difference between Kepler and Daloopa?

Kepler is the platform around the numbers: statements assembled from each filer's XBRL with every cell cited, your firm's ontology and definitions, every calculation run in code, agents that update the work when companies file, and every surface your analysts use. Daloopa supplies source-linked historical data points on 6,000+ public companies, delivered to Excel, an API and AI assistants for another tool to reason over.

Does Kepler give source-linked historicals like Daloopa?

Yes. Every statement line, KPI and segment figure in Kepler opens the exact page, line or cell it came from, and every derived number opens its formula and inputs. Statements come from the filer's own XBRL with no model in the path, and KPIs from MD&A tables, footnotes and earnings decks are read at ingest with their scale and unit. Coverage spans 14,000+ companies in 120+ countries.

How do Kepler and Daloopa handle restatements?

Daloopa's documentation says it keeps both the originally reported and the restated value for a period, flagged in its data. Kepler returns either view on any date and rebuilds each company's history so a line keeps one meaning across years; in our research, 194 of 200 companies reshaped a statement in ways XBRL tags can't follow. Recasts are flagged with both versions shown.

How much does Daloopa cost?

Daloopa doesn't publish prices. Its plans page lists a free tier with up to three data sheets, Core (data sheets, the Excel add-in, Scout and MCP access with monthly limits), Premium (Core plus API access) and a separate Fundamentals API plan, each through sales. Scout usage is metered in monthly credits set by the plan.

Can I use Daloopa or Kepler inside Claude or ChatGPT?

Both. Daloopa's MCP connector puts its data inside Claude, ChatGPT and Microsoft 365 Copilot, where the assistant reasons over the data points. Kepler runs inside Claude and ChatGPT as well, and the numbers still come from Kepler's engine: computed in code, cited to the line and on your firm's definitions.