Financial intelligence built for AI agents. Auditable and approved by humans.
Build your analysis, valuation models, reports, and verify every number in minutes. Share them with your team or clients.
Total revenue, as reported
$416.16B
Works in your browser, from Python, or inside the AI you already use.
- No look-ahead bias
- You approve anything irreversible
Two datasets: financial fundamentals and smart-money.
Financial fundamentals
111M+ facts · 19,000+ companies · 1993–present
Free tier and up
Smart-money signals
78M+ rows · Forms 3/4/5/144 + 13F/13D/13G
Institutional
The model never mints a number.
Typed tools return SEC-filed values; the model only arranges the words. Same inputs, same output — and every figure is one click from its filing.
A tool returns the number.
META · op_income · FY2024
$69,380,000,000
There is no free-text digit to invent — the model selects a fact, it doesn't author one.
Same inputs, same output.
temperature 0 · point-in-time snapshots · deterministic tools. Research your firm can reproduce years later.
One click to the filing.
No re-running, no trust-me. An auditor clicks through and checks the figure against the original 10-K.
We didn't ask the model to stop hallucinating. We removed its ability to be the source of a digit.
Read the full methodologyMost numbers that change, change quietly.
We kept every version of every figure filed since 1993 — including for companies that no longer exist. That makes a question answerable which the latest filing alone cannot answer: when a company revises a number it already reported, is anyone told?
- 497 / 500
- have revised something
- S&P 500 companies with at least one revision we detect since 1993
- 23
- ever filed an Item 4.02
- the formal 8-K notice telling investors not to rely on past financials
- 94.5%
- arrived in a routine filing
- revisions that landed in an ordinary 10-Q or 10-K, with no amendment in our filing index
This is a finding about disclosure, not wrongdoing. A revision is often an accounting-rule change — ASC 606, ASC 842 — restating prior periods with nobody at fault, and the absence of an amendment in our index is not by itself evidence of anything. We show you both filings and the exact figure that moved. You draw the conclusion.
Most “historical” data is a lie told in hindsight.
It silently overwrites with restatements and quietly drops the companies that failed. Both errors inflate every backtest. Here is what ours does instead — visibly.
| Metric | as reported | restated |
|---|---|---|
| Revenue | 10,388 | 10,388 |
| Operating income | 2,114 | 1,902 |
| Net income | 1,640 | 1,431 |
| Diluted EPS | 3.05 | 2.66 |
verify_fact_lineage() → both versions retained · each keyed to its filing
- ENRNbankrupt · 2001
- LEHbankrupt · 2008
- WAMUseized · 2008
- BBBYbankrupt · 2023
- SIVBfailed · 2023
- FTXcollapsed · 2022
get_pit_universe(date) → the index as it stood, not as it survived
One token. Three channels.
One subscription, three ways in — an AI agent, a browser workspace, or your own code. Same data, same lineage, same persisted state underneath: save a thesis in Claude, see it in Cursor.
The agent you already use, now a cited equity analyst.
Any MCP client — Claude, ChatGPT, Cursor, Windsurf or VS Code — wired to 118 typed tools and 39 expert workflows. Works the moment you connect, pay per call to start, every result cited.
Connect your AI agentCalling Valuein MCP…
Your agents run while you sleep. You approve what matters.
A browser research desk with your own LLM key — sealed for 24 hours, never trained on. Agent teams re-run your workflows on every new filing; every high-impact action waits for your approval. Human-on-the-loop, not human-out-of-the-loop.
Try the WorkspaceDrafting the brief into the document beside you.
reading the FY2024 10-K…
NVDA — Q1 research brief
$130.50B
└─ 10-K · fact_id 4c1f…9ab2
The backtest that can't cheat.
pip install, one token, 111M+ standardized facts through DuckDB — point-in-time, survivorship-free, zero look-ahead. The raw truth in your own code, not a black box.
Read the SDK docsfrom valuein_sdk import ValueinClient # Point-in-time SEC fundamentals — sample tier, no API keywith ValueinClient() as client: df = client.run_template( "fundamentals_by_ticker", ticker="META", ) print(df.head())One capability chain. Every desk's outcome.
The same platform — codified workflows, batch runs, a shared library, automations — lands differently on every desk. Pick yours.
Financial analysts
Spread any 10-K in seconds, cited to the filing.
Portfolio managers
Run the whole book — an analytical partner, not an autopilot.
Quants
Backtest 30 years without survivorship bias.
Researchers
Reproducible, restatement-aware data for published work.
Developers & AI builders
Wire SEC data your agent can't fabricate.
Data engineers
One clean feed — no scraper farm, no XBRL wrangling.
Independent investors
Institutional-grade research on a personal budget.
Research & finance leaders
AI you can put in front of a risk committee.
Start on the house. Scale when the work does.
Begin on the S&P 500 with no card, then open the full universe and smart-money signals as the work grows. Paid plans are priced per seat — per person, never per firm.
Free — S&P 500
$0no card
Begin on the S&P 500, 1993–present.
- Full S&P 500 history
- MCP, SDK & Workspace
- One token across channels
Pro
$49/mo per seat
The full universe for analysts & developers.
- 19,000+ companies, incl. delisted
- 15-year point-in-time window
- 118 tools + 39 workflows + SDK
Institutional
$499/mo per seat
Full history, smart-money, redistribution.
- 1993–present, all amendments
- Insider + 13F smart-money data
- Webhooks, SLA & redistribution
Add one connector. Get a research floor.
Free tier, no card — real filings in under 15 seconds. Connect an agent or open the Workspace to start.