Legal AI for solo attorneys and small firms.

It is 10:04 p.m. on a Tuesday. A solo practitioner I know — employment law, Midwest, one paralegal part-time — is drafting a motion for summary judgment that is due tomorrow. She has four tabs open. One for Westlaw, at $300 a month. One for a docket-monitoring tool, another $150. One for a general consumer AI chat product, because the drafting suggestions are faster than typing. One for her local federal rules. She is doing what most solo attorneys do every week in this country: holding together a stack of tools that was never designed for her, paying more than BigLaw pays per seat, and double-checking every line because one of those tabs will cheerfully invent a case.

This is the reason legal AI for solo attorneys has been, for three years now, a broken category. The good tools were priced for firms with a head of knowledge management. The cheap tools were general-purpose products dressed up with a legal veneer. There was no middle. Not anymore. This piece is the guide I wish someone had handed her at 10:04 that night.

What solo attorneys actually need from legal AI

Start from the job, not the feature list. A solo or a two-to-ten-attorney firm needs four things from a legal AI platform, and everything else is noise.

Research parity with BigLaw tools. When opposing counsel at a 300-attorney firm runs a search, they are hitting every U.S. court opinion back to the 1700s, every federal statute, every state code. A solo should hit the same corpus. Aewita covers every U.S. court opinion from 1700 to today, federal statutes, and statutes for all 50 states plus D.C. If your legal AI cuts corners on coverage to hit a price point, you are not actually saving money — you are shipping a motion with a gap.

Drafting support that does not hallucinate. A general consumer AI will happily suggest a case that does not exist. That is not a solo-friendly tool. That is a sanctions risk. In our internal testing, Aewita observed zero hallucinated outputs across 800 consecutive queries — statistically, a rough upper bound under 0.3% at 95% confidence. Every citation is independently verified before it reaches you. You are still the lawyer. You still read. But you are not fighting the tool.

No enterprise contract. A solo should not need a master services agreement, a data processing addendum, a six-month procurement cycle, and a committee to approve a research subscription. A credit card should be enough.

Cancel anytime. Your practice changes. You might pick up a big case that justifies more tools, or lose a client and need to trim. A legal AI that locks you into a three-year commitment is not built for a solo.

The pricing reality: enterprise AI versus $99 a month

Enterprise legal AI keeps solos and small firms out by design. The lock-outs are not subtle:

• Seat minimums. Five seats. Ten seats. Twenty-five. If you are a two-attorney firm, you are paying for seats you do not have.
• Annual commits. Many enterprise legal AI products will not sell month-to-month. The list price, if you can find one, assumes a year up front.
• Onboarding fees. A one-time five-figure charge for "implementation" that, for a small firm, mostly means a login.
• Committee approval. A vendor sales cycle built for a firm with a CIO, a director of KM, and a procurement team. That sales cycle does not end with a $99 monthly charge. It ends with a multi-year contract you did not need.

Aewita’s list price is $99 per month or $720 per year, with a 14-day free trial and no seat minimum. That is the whole quote. No separate line items. No onboarding fee. No committee. The math problem for a solo is small enough that the decision is an hour, not a quarter.

Tool consolidation: the stack math for small firms

Walk through what a typical small-firm technology stack actually costs when you add up what most solos I have talked to are paying each month. Numbers vary by practice area and vendor, but a representative stack looks like this:

• Premium legal research subscription: $200–$400 per month per attorney.
• A second research tool, because the first one does not cover a specialty jurisdiction well: another $75–$150.
• A standalone AI drafting assistant: $30–$50.
• A citation checker: $30–$75.
• A general-purpose AI chatbot, because everyone is using one: $20.

That is $350 to $700 a month per attorney, across four or five logins, each with its own billing cycle and its own security posture to audit. For a solo, the bottom of that range is already north of Aewita’s annual price — and you still do not have a tool that can draft a motion, check its own citations, and sit inside a privilege-preserving architecture.

One Aewita subscription replaces most of that stack. You still want access to a primary-source research service if that is how you work — and many solos keep theirs running alongside Aewita during the first few months of a trial, which is the right instinct. But the drafting, the citation verification, and the coverage gaps all collapse into a single subscription.

Solo doesn’t mean second-class. The same legal corpus, the same AI, at a price that fits.

What to skip when you evaluate legal AI for a small firm

Most solos overpay for things they do not need. Three categories to be specific about:

Do not pay for seat-minimum tools. If the sales rep cannot quote you a single-seat price on the first call, the product is not built for you. Move on. The same applies to "enterprise" pricing pages that require a contact-us form to see a number.

Do not pay for enterprise DMS integrations if you do not have a DMS. Plenty of legal AI platforms charge extra for their integration modules with iManage, NetDocuments, or SharePoint. If your firm runs on a shared Dropbox folder and a well-organized desktop, you do not need those modules, and you should not subsidize them. When you do eventually adopt a DMS, Aewita connects over the Model Context Protocol without a migration.

Do not pay for "AI features" layered on top of research tools you already use. Several incumbents now charge a premium add-on for an AI-assist inside a product you were already licensing. Look at the add-on carefully. If it is a re-skin of a general-purpose model bolted onto the same database you already had, you are paying twice for overlap. You want a legal AI that was designed end-to-end, not retrofitted.

How to start: a 14-day trial plan

Here is the concrete plan I recommend to solos and small-firm partners who sign up for the 14-day free trial. Three queries in the first week. Each one tells you whether the tool is the real thing.

Query one: the hardest research question on your desk right now. Not a test question. The real one. The one you have been circling for three days. Run it. Read the citations. Open two of them and verify they say what the output claims they say. This tells you whether Aewita’s research layer meets your standard.

Query two: a drafting task you normally do by hand. A response brief section. An MSA indemnification clause. A demand letter. Ask Aewita to draft it, grounded in the filings or the fact pattern you paste in. Read the draft. Check the citations. Edit it the way you would edit an associate’s. This tells you whether the drafting is worth the typing-speed it saves.

Query three: a jurisdiction edge case. An obscure state-court procedural rule. A recent appellate opinion you happen to already know the answer to. Run it, and see whether Aewita’s answer matches ground truth. This tells you whether coverage is real or a marketing claim.

If all three clear the bar, you have replaced two or three line items in your stack with a tool that fits your practice. If any of them fall short, cancel before day fourteen and move on — that is the point of the trial.

Why solos are actually the right fit for legal AI

The standard story is that BigLaw gets AI first and solos get it last. The truth is closer to the opposite. Solo and small-firm practitioners make up a significant share of the U.S. legal profession — by some tallies, around 40% of law firms are solos and more than three-quarters have fewer than six attorneys. They do more of their own research. They write more of their own drafts. They have no knowledge-management team to slow-walk technology through a committee. When a tool is priced fairly and it works, a solo can adopt it on a Tuesday and be using it by Thursday.

That is the audience this product was built for. The AmLaw 200 will catch up when they catch up. You do not have to wait.

If you have been watching the legal AI category from the sidelines because every product you looked at required a procurement cycle — you are in the right place. The 14-day trial is a real trial, with real product access. Bring your hardest question. See where it lands. Compare it against the stack you are already paying for. If it does not earn its keep, you owe us nothing. That is the deal I am comfortable making, because we built it, we host it, and every citation is independently verified before it reaches you.