See an AI-native platform
built for solo & small firms
Registration is open for solo lawyers and small firms.
| Referent | Paxton AI | |
|---|---|---|
| Category | Operations: runs the firm | Legal research & drafting (affordable) |
| The job it does | Intake, matters, billing prep, follow-ups | First-pass research and drafting |
| What the AI does | Runs operations; you approve | Answers research questions and drafts |
| Data it works from | Your firm's live matters | Legal sources and uploaded documents |
| Pricing | Free plan; paid plans, AI usage included | From ~$99/seat/mo |
| Replaces the other? | No, complementary | No, complementary |
| Best for | Running a small firm, AI-native | Budget-conscious research and drafting |
How we compare: Paxton AI details are from public vendor sources as of June 2026; Referent's reflect current private-beta capabilities. Referent is our product, so we note where Paxton AI wins below, and you should verify current plans on each vendor's site.
Across US law firms, only about 3 of 8 working hours are billable. The rest is admin Referent's AI agents run, with you approving. See the data.
Try for freeReferent and Paxton AI get compared, but they are not competitors. They do different jobs. Referent is an AI-native practice management platform built on AI agents for law firms: its agents run intake, matters, billing prep, and follow-ups while the lawyer approves. Paxton AI is an affordable legal-research-and-drafting assistant aimed at solo and small firms. The honest verdict is that this is not either/or. For a small firm, Paxton (research) plus Referent (operations) is a low-cost, full AI stack.
Referent runs the firm; Paxton answers the law. Referent is the operating layer. It moves operations forward from your live matters and stages them for approval. Paxton AI does first-pass research and drafting at an accessible price. Both use AI, for different jobs.
That operations load is where most of a lawyer’s day actually goes. Solo and small-firm data show only about 3 of every 8 hours are billable, with admin eating the rest. Research AI like Paxton makes the billable hours faster. Referent is the law firm automation software that shrinks the non-billable ones.
Paxton AI is a solid, affordable entry into legal AI:
For the research category and where Referent sits alongside it, see Paxton AI alternatives.
Moving off Paxton AI is not a rip-and-replace project. Referent's white-glove onboarding connects your Gmail, Google Calendar, and Google Drive and sets up your matters, so a working AI-native baseline is live in days, not months, with the lawyer approving every client-facing action from day one. Plan a separate tool for anything you rely on Paxton AI for that Referent does not cover, such as built-in accounting. Referent is in private beta, so onboarding is hands-on and personal.
Not exactly. They do different jobs. Paxton AI does affordable legal research and drafting; Referent runs a firm's operations (intake, matters, billing, follow-ups) with the lawyer approving. Referent does not do legal research. Many small firms use one of each.
Referent is an AI-native practice management platform whose agents run the firm's operations. Paxton AI is an affordable legal-research-and-drafting assistant. One runs the operation; the other works on the law. They are complementary.
No. Referent runs operations (intake, matters, billing prep, follow-ups) with AI agents and lawyer approval. For affordable research and drafting, pair it with Paxton AI.
For most firms it is not a choice between the two. If you need affordable research and drafting, choose Paxton AI. If you want AI to run your firm's operations, that is Referent. Paxton plus Referent is a low-cost full AI stack for a small firm.
Paxton AI starts around $99 per seat per month; Referent starts free, with paid plans and AI usage included. They are billed for different jobs (research versus operations), so together they form one affordable stack rather than a like-for-like choice.
Registration is open for solo lawyers and small firms.
Set up in minutes
Workspace built from your inbox
No data training