Everything you need to know about Legatus AI. Can't find your answer? Email us at hello@legatus.legal.
You need three things: a Legatus AI account, a Google AI Studio API key (free), and a probate case to work on. Setup takes about 10 minutes. You enter your firm's details once in Firm Settings and they autofill into every case and every pillar going forward.
Optional but recommended: connect Plaid for automatic bank data, and Clio or Estateably if your firm uses those platforms.
Go to aistudio.google.com, sign in with your Google account, and click "Get API key." It's free with generous usage limits — enough to run hundreds of cases per month before any cost kicks in.
Alternatively, if your firm prefers Anthropic's Claude, you can use an Anthropic API key instead. Both are supported in Firm Settings.
No — Legatus AI works alongside your existing software. It handles the four administrative tasks that Clio and Estateably don't: finding hidden assets, managing creditor claims, writing beneficiary updates, and cancelling subscriptions.
Think of it as a specialized tool that fills the gap your practice management software leaves behind.
About 2 minutes. You enter the deceased's name, date of death, case number, and jurisdiction once at the top. Those details autofill into all four pillars automatically. Then each pillar has its own focused form for the specific data it needs — bank statements, creditor claims, beneficiary details, or subscription lists.
Client estate data is never stored on Legatus AI servers. When you run a pillar, the data you enter is transmitted directly to the AI provider (Google or Anthropic) to generate the output, and that's it. Nothing is logged or retained on our end between sessions.
All connections use HTTPS encryption. API keys are transmitted securely and never stored by us.
Yes — and we recommend it. In Firm Settings, enter your own Google AI or Anthropic API key. When you do, your case data goes directly from your browser to your AI provider account — it never touches our platform at all. Your firm controls the entire data flow.
That determination is yours to make as a licensed attorney based on your jurisdiction's rules. We recommend reviewing your state bar's guidance on cloud-based legal tools and AI in legal practice.
What we can tell you: data is not retained on our servers, you control which AI provider processes your data, and you can use your own API keys for complete isolation. Many firms find this satisfactory — but your professional judgment governs.
No. Legatus AI is an administrative tool. It does not practice law, does not give legal advice, and does not create any attorney-client relationship between Legatus AI and any party.
Every output — asset reports, creditor letters, beneficiary updates, cancellation notices — is a draft that requires attorney review and approval before use. The supervising attorney is solely responsible for all outputs.
Nothing goes out without your review. Every pillar produces drafts — not final documents. The attorney reads, edits if needed, and approves before anything is sent to a creditor, beneficiary, or service provider.
That said, AI outputs are generally accurate for the tasks Legatus is built for. The asset discovery and creditor analysis are based on the data you provide, and the letters follow standard probate communication formats. Treat every output the way you'd treat work from a very capable paralegal — review before sending.
Typically 10–30 seconds per pillar. If you're running it for the first time after the server has been inactive, it may take up to 60 seconds while the server wakes up — this is normal and only happens on the free hosting tier.
In the current version, you paste bank statement text into the Asset Discovery pillar and the AI identifies signals — a transfer to Coinbase, a recurring payment to a crypto exchange, email receipt language indicating an account. It identifies these from the data you provide rather than connecting directly to exchanges.
Automatic crypto account detection through direct API connections is on our Phase 2 roadmap.
Currently Legatus AI works alongside Clio — you run both tools, with Legatus handling the tasks Clio doesn't cover. Full Clio integration (auto-importing contacts, matters, and case files) is on our roadmap for Enterprise customers.
If Clio integration is important for your firm, let us know — it directly influences our build priority. Email hello@legatus.legal.
In Firm Settings, enter your Plaid Client ID and Secret Key under Banking integrations. Legatus AI will then use your Plaid connection when running Asset Discovery and Estate Preservation.
Note: the current version uses your Plaid credentials to inform the AI's analysis. The full Plaid Link flow — where executors log into their bank through Plaid's secure interface — is coming in Phase 2.
Yes — in Firm Settings under "Custom APIs & subscriptions," you can add any API key your firm uses and specify which pillars should use it. These credentials are stored in your browser session and sent with every relevant AI request.
Plans are priced per seat per month:
No setup fees. No long-term contracts. Cancel anytime.
The live demo at legatus.legal/dashboard.html lets you explore the full product with the Henderson Estate sample case. You can run the AI on real data during an onboarding call with our team before committing to a subscription.
Yes — no long-term contracts, no cancellation fees. Cancel by emailing hello@legatus.legal and your subscription ends at the close of the current billing period.
By default, AI usage is included in your Legatus subscription — we absorb the cost. If you prefer to use your own API key (for data isolation or cost control), you can enter it in Firm Settings and AI usage is billed directly to your AI provider account instead.
Use the "⚠ Report an issue" button at the bottom of the dashboard sidebar. Describe what happened and which pillar is affected — we'll respond within one business day.
For urgent issues, email hello@legatus.legal or call 212-518-8829.
If the server hasn't been used recently it may take up to 60 seconds to wake up on the first request — this is normal. The dashboard shows a message when this is happening. Simply wait and the response will come.
If it's been more than 2 minutes with no response, refresh the page and try again. If the problem persists, use the Report an Issue button.
Three ways:
We respond within one business day, often same day.