Compare

Zefort and Lexnus both check contracts against your own rules. They disagree about what a rule is.

Zefort's Playbooks let you write a policy rule as a prompt and have AI check contracts against it. Lexnus extracts the value a rule names, then evaluates it against a playbook you have published and frozen. Same goal, different machinery — and the difference shows up the day you have to prove which rule judged a contract.

The difference in one line

Zefort: Prompt-evaluated rulesYou describe what the AI should look for. It reports compliant, not compliant, unclear or no information, with references back into the contract text.
vs
Lexnus: Extract, then evaluateThe AI extracts the value your rule names. The rule compares it. The answer is pass or fail against a playbook version that cannot change underneath it — or the escalation that rule specifies.

Both have playbooks. They are not the same object.

Zefort ships a feature called Playbooks. You define internal policies or legal requirements, and contracts are checked against them automatically, with the results pointing back at the clause text that triggered them. If you have read somewhere that a contract repository cannot enforce policy, that is out of date, and this page is not going to pretend otherwise.

The difference is what a rule is made of. In Zefort a rule is a prompt: you write a description of what the AI should look for, and the model interprets it. A result can come back compliant, not compliant, unclear, or no information. Zefort's own documentation is candid about the consequence — it recommends treating playbook results as a helpful guide rather than a final legal determination, and spot-checking the ones that come back unclear.

Lexnus splits the job in two. The model reads the contract and extracts the values your rules reference — the cap, the notice period, the governing law. Then a published rule compares that value against your standard. The model does the reading, which it is good at. The rule does the judging, which it is not. That boundary is the reason someone in sales can act on the output without a lawyer translating first. Where a rule is not meant to be settled automatically it carries an escalation type, and the contract goes to a named person — which is itself a decision a lawyer made once and froze in the version.

The second difference is what happens when the standard changes. Edit a rule in Zefort and the contracts already evaluated against it move to "not evaluated" until they are scanned again; the rule updates in place. In Lexnus, publishing freezes the playbook. An edit creates a new version, the old one stays exactly as it was, and every verdict already issued remains attached to the version that issued it. A year later you can still show which rule judged the contract and what it said at the time.

The third difference is where the playbook can go. A Lexnus playbook is delivered as well as stored: an agent over MCP can pull every rule in a published version, with its conditions and escalation type attached, and check a set of terms against them before a contract exists. The policy is an object another system can hold, not only a setting inside ours.

Side by side

 AI playbook reviewLexnus
What a rule isA prompt describing what to look forA comparison against an extracted value
What reads the contractThe modelThe model
What decides the outcomeThe modelThe rule
What a result can beCompliant, not compliant, unclear, no informationPass or fail on each rule
What sends it to a personAn unclear result, which their documentation suggests spot-checkingAn escalation type on the rule, set when the version was published
When you edit a ruleEarlier results are invalidated and re-scannedA new version is published; earlier verdicts keep theirs
What the audit trail provesWhat the current rules find todayWhich policy version judged it, and what it said

This page describes an architectural difference, not a feature scorecard, and it was checked against Zefort's own published documentation in September 2026. Products change; for what any vendor ships today, read their documentation rather than ours.

When Zefort is the right answer

You need the archive job done properly: thousands of signed agreements, extracted and tagged, with obligations and renewal dates surfaced before they surprise you. Zefort is built around that and does it well. Lexnus is not competing for it.

It is also the right answer when "unclear, have someone look at it" is a perfectly good outcome for your rules. Plenty of policy is genuinely interpretive, and a prompt-evaluated review handles interpretive questions that a strict comparison cannot express at all.

Lexnus is the better fit when the opposite is true: when a rule has a right answer, the same contract must get the same verdict every time, and somebody will eventually ask you to prove which version of the policy produced it.

Questions

Does Zefort not already do this?

Part of it, yes. Zefort checks contracts against rules you define, and if your question is only "can this product hold me to my own policy at all", both answer yes.

The difference is in the two places above: what decides the outcome, and what survives a change to the rule. A prompt evaluated by a model returns the model's reading. A comparison against an extracted value returns pass or fail, and escalates where a rule says it should. And a rule that updates in place cannot tell you later what it said before you changed it.

Why does rule versioning matter?

Because the question usually arrives late. A deal is disputed, an auditor asks how a term got approved, or a regulator wants to see the control that was actually in force in March.

If rules update in place, the honest answer is what the policy says now, not what it said then. Lexnus keeps published playbooks immutable, so a verdict stays traceable to the frozen policy state that produced it.

Can we use both?

Yes, and the split is clean. Zefort holds the record of what you signed and answers questions across it. Lexnus governs what you sign next and records which rule version allowed it.

Lexnus also reads signed contracts, but for a different reason: to draft your first playbook from the positions you have already been taking. If your archive is in order, that first hour goes faster.

See it on your own contract.

Upload one contract you have already signed. Lexnus drafts a playbook from it, and you can watch your own standard get enforced. Free trial, no card.