Law Firm Intake Software 2026: Clio vs $8k Owned AI

A law firm AI intake system that writes to the CRM is not a complicated build. But it breaks down in predictable ways when the workflow is not designed around two non-negotiable rules: capture facts, never give legal advice.

Short answer: Law firm intake software should capture the call, write clean facts into Clio, Lawmatics, MyCase, or your current CRM, and route judgment calls to a human. As of August 2026, rented AI receptionist plans run from a free 25-call tier or $165/month entry minutes to $500/month-plus. My owned AI Receptionist build is $8,000 one-time.

The rest of this covers the exact workflow map, what each major CRM costs this month, where the ownership math changes, and when to hold off on deploying this at all.

Buyer question	Best fit	Current pricing signal (Aug 2026)
“law firm intake software”	Clio Grow, Lawmatics, MyCase	CRM subscription plus intake forms
”automated legal intake”	AI receptionist plus CRM write-back	$165–$500/mo rented, or $8,000 owned
”AI CRM for attorneys”	Usually not a new CRM	Put AI in front of the existing system
”legal CRM assistant”	Call capture, summaries, routing	Value depends on inquiry volume and handoff rules
”AI intake platform law firm”	Smith.ai, Abby, or owned build	Metered SaaS vs one-time workflow you keep

How does legal intake automation work, step by step?#

Legal intake automation is a five-step workflow: classify the caller, collect structured facts, capture conflict-review names, write the CRM note, then route by the firm’s rules. The AI follows the script and stops before legal judgment.

Each step is deterministic — the AI follows your rules, not its own judgment.

Trigger: Inbound call or web form submission from a prospective client.

Step 1 — Caller classification. The AI asks whether the caller is a new prospective client, an existing client, opposing counsel, court staff, or something else. This matters because law firm phones mix new leads with client service calls, vendor calls, and opposing counsel calls. Treating every call as a new intake creates CRM noise and wastes paralegal time.

Step 2 — Structured intake collection. For a new prospective client, the AI collects: full name, callback number, email, matter type in the caller’s own words, jurisdiction, date of incident or issue, any known court dates or response deadlines, opposing party names for conflict review, whether they have already hired an attorney, and how they found the firm. Practice-area-specific fields extend this base — a personal injury intake looks different from a family law intake.

Step 3 — Conflict flag. The AI does not run a conflict check. It captures opposing party names and flags them for your review. Safer language: “I’m collecting names for the firm’s conflict review process.” Do not build an AI that tells callers a conflict is cleared unless you have explicitly approved that automated workflow.

Step 4 — CRM write. The AI sends a formatted intake note to your CRM’s lead or matter record via webhook or API. A useful CRM note reads like a human paralegal wrote it: “New PI intake. Caller: Marcus T. Callback: 555-0177. Email provided. Jurisdiction: Cook County, IL. Incident: car accident, 3/14/2026. No known court date. Opposing party: unknown at this time. Has not hired counsel. Found firm via Google. Requesting consult this week.”

A bad note: “Caller wants help with a car accident.”

That second version requires a second call to fill in the blanks. The whole point of intake automation is eliminating that second call.

Step 5 — Routing and escalation. Qualified lead gets a consult offer. Wrong practice area gets a polite referral script. Urgent call — arrest, imminent court date, safety issue — gets an immediate human alert. The AI does not decide what urgent means. You define that in the routing rules.

This is the workflow behind Legal Intake AI Receptionist.

Which law firm intake software should a small firm use?#

For most solo and small firms, the practical answer is not “buy a new AI CRM.” Keep the system of record that already holds matters and contacts, then add a controlled intake layer that writes into it.

Three platforms come up most often for solo and small firm intake.

Clio Grow is Clio’s intake and CRM product, sitting beside Clio Manage. According to Clio’s pricing page, as of August 2026 Manage plans are named Starter, Core, Signature, and Elite, starting at $49 per user per month. Grow is included on Elite and Elite Work and sold as an add-on on Core and Signature. Older writeups that say Grow is a flat $59 add-on on “Essentials” or “included on Advanced” are describing a plan map Clio no longer publishes. Get a current quote.

Lawmatics is a purpose-built legal CRM with intake automation baked in. According to Lawmatics’ pricing page, as of August 2026 Essential, Premium, and Enterprise are custom-quote plans with a 3-user minimum (Enterprise lists a 5-user minimum on the compare table). Essential still lists 500 contacts, client intake, lead management, calendaring, forms, e-signature, SMS, and tasks, with 10 automations on Essential versus unlimited on Premium. QualifyAI / Merlin is an add-on with credit or outcome-based pricing — Lawmatics no longer publishes a simple public dollar figure on the old Merlin help article, which now 404s. Ask sales for the current credit pack.

MyCase now runs $50 / $100 / $130 per user per month for Basic, Pro, and Advanced when billed annually, or $60 / $120 / $150 on month-to-month, according to MyCase’s official help article dated June 1, 2026. Pro is the default intake-capable tier. Some third-party roundups still quote older $39 / $89 / $109 cards; I am using MyCase’s own numbers.

For AI receptionist write-back, all three support webhook or API-based intake note creation. The integration depth varies by tool, but a properly deployed AI system can write a structured CRM record to any of them.

One note for anyone searching for an “AI CRM for attorneys and legal practices”: you almost never need a new CRM. The AI layer belongs in front of the system you already run — answering the phone, collecting the facts, writing the note. For a solo attorney who’d rather drive that capture from a chat app, the phone-first Telegram intake workflow is the same idea sized for one operator. Swapping CRMs mid-practice adds migration pain without fixing the actual gap, which is at the phone, not the database.

What does this cost, and what do you actually own?#

The cost question is not only monthly price. It is whether the firm owns the intake script, CRM write-back logic, routing rules, recordings policy, and escalation behavior after the vendor invoice stops.

SaaS AI receptionists for law firms can look cheap at the entry tier, but the cap matters. According to Smith.ai’s AI Receptionist pricing, as of August 2026 the free tier includes 25 calls/month with $3.00/call after that. Pro starts at $150/month for 75 calls ($2.00 in-plan, $2.50 extra), then $270/month for 150 calls and $500/month for 300 calls. Enterprise starts at $500/month for 300 calls.

Abby no longer publishes a separate $99 AI-only starter. According to Abby Connect’s pricing page, as of August 2026 usage is Abby Minutes on one card: Starter $165/month for 50 Abby Minutes, Essential $329 for 100, Professional $599 for 200, Growth $1,380 for 500. Human answering uses one Abby Minute per minute. AI answering uses half a minute, so Starter stretches to 100 AI minutes if every call stays on AI.

A live human answering service is a different bill. According to Ruby’s plans page, third-party recaps that checked ruby.com in late August 2026 still list $250/month for 50 receptionist minutes up to $1,725/month for 500 minutes. That is the premium live-voice lane, not the AI CRM write-back lane.

A firm taking 30-plus new inquiries a month on top of existing-client calls can clear entry limits quickly. At 90 Smith.ai AI calls on the 75-call Pro plan, the published math is $150 plus 15 extra calls at $2.50 — about $187.50/month before human backup or extra services. For the long-run cost of renting one of these versus owning the call flow outright, see the Smith.ai alternative breakdown.

	Smith.ai AI Pro	Abby Starter (AI mix)	MyCase Pro	One-time deployment
Monthly cost	$150/mo for 75 calls	$165/mo for 50 Abby Minutes	$100/user/mo annual	$0/mo after build
12-month cost	$1,800 plus overages	$1,980	$1,200/user	$8,000 one-time
24-month cost	$3,600 plus overages	$3,960	$2,400/user	$8,000
36-month cost	$5,400 plus overages	$5,940	$3,600/user	$8,000
Owns workflow?	No	No	No	Yes
Custom CRM write?	Native integrations, vendor-owned	Integrations, vendor-owned	Native CRM	Built to spec

At 36 months, a $150/month receptionist is still cheaper than an $8,000 build if volume stays inside the tier and the generic workflow fits. That is the honest math. The owned deployment starts winning when the firm needs custom CRM fields, stricter legal boundaries, no per-call meter, and the same workflow running for years.

One caveat on the table: the Smith.ai and Abby figures are entry-to-mid tiers. A firm with real inquiry volume lands a tier or two up, and MyCase or Lawmatics is still a CRM cost, not the whole phone-answering layer. The owned build includes the intake call flow, CRM write-back logic, routing rules, and escalation script — built around how your firm actually operates.

For the full pricing model comparison, see the AI receptionist pricing breakdown.

Is this cheaper than an intake coordinator?#

Compared with a full-time intake coordinator, AI is cheaper on coverage and worse on judgment. Use it for after-hours capture, CRM notes, and routing; keep legal evaluation, sensitive callbacks, and final acceptance decisions with people.

Do not budget an intake hire off a round $40k salary plus a made-up benefits add-on. According to eNZeti’s 2026 intake coordinator salary guide, median full-time U.S. law-firm intake coordinator pay sits between $42,000 and $52,000, with fully loaded cost often $55,000–$90,000 once benefits, taxes, and training are in. According to Smith.ai’s June 2026 intake-specialist pay roundup, small and mid-size firms commonly budget $35,000–$60,000 in salary, and all-in cost routinely lands above $70,000. That person still works business hours. No weekends, no 8pm calls, no coverage during depositions.

An AI intake system that writes to the CRM does not take PTO. The AI vs. in-house staff cost comparison runs the year-one and year-three numbers in detail.

For a 3–5 attorney firm, the issue is usually not that the coordinator is doing a bad job. It is that calls come in at 8pm on weeknights and Saturday mornings when prospects have time to search for attorneys. That is when the AI carries the load.

What should happen next in law firm lead capture automation?#

After the call, the next step is a complete CRM record plus a human decision — not a voicemail, not a chat log, and not a second fact-finding call. Law firm lead capture automation fails when the AI captures the lead and then dumps it into email.

This is the merged “what should happen next” path I see firms skip:

1. New lead record exists before the caller hangs up. Name, callback, matter type, jurisdiction, dates, opposing-party names, source.
2. Conflict names sit in a review queue, not in a “cleared” field.
3. Qualified leads get a consult slot or a same-day callback task assigned to a named person.
4. Wrong-fit callers get a referral script, not a fake “we’ll take it.”
5. Urgent language triggers a live alert, not a morning digest.

If your stack stops at “we got an email,” you do not have automation. You have a nicer voicemail. A legal CRM assistant is useful only when those five handoffs are explicit.

How does the AI avoid giving legal advice?#

The boundary is a script decision, not an AI decision. The system must be told exactly what to collect, what to refuse, and when to hand the call to a person.

The AI must be told — in the system prompt and the call flow design — exactly what it is and is not allowed to say. That means:

• It can collect facts.
• It can confirm a consult will follow.
• It can say it is collecting information for the firm’s review.
• It cannot evaluate whether the caller has a case.
• It cannot explain legal rights or tell the caller what their options are.
• It cannot tell the caller a conflict check is cleared unless that is an explicitly approved automated step.

That red-light list is worth getting exactly right — I walk through the full set of things an AI intake agent must never do at a law firm and why each one is a licensed-attorney decision.

ABA Formal Opinion 512, issued July 29, 2024, addresses attorneys’ obligations when using generative AI, including competence, confidentiality, communication, and supervision duties. For intake, the practical translation is: the AI collects data under the attorney’s supervision. The attorney reviews and takes responsibility for what happens next. The AI does not operate independently.

That line is straightforward to maintain if it is built in from the start. It breaks when a system is deployed with vague instructions and someone expects the AI to improvise on hard legal questions.

For a deeper look at what the AI can and cannot capture safely, legal intake AI receptionist — what to capture covers the field-by-field breakdown.

When is this the wrong move?#

Do not automate legal intake until the firm can write down its accepted matter types, conflict-name rules, urgency triggers, CRM destination, and human review path. Automation follows the firm’s process; it does not create one safely.

There are real cases where I would tell a firm to wait.

Your intake questions change every call. If every new matter genuinely requires attorney judgment before you know what to ask, scripting the AI is harder than the savings justify. Intake automation works best when the information you need is consistent across callers.

You are under 30 inquiries a month. A low-tier SaaS tool or a simple missed-call text-back covers the need at that volume. A custom build is overbuilt until inquiry volume and case value justify it.

You have not mapped your current intake flow. This is the most common failure mode I see. If you cannot describe what questions your intake process asks in what order, the AI cannot be built around it. Automation follows process — it does not invent one.

Your jurisdiction has bar rules about initial client contact. Some state bars regulate what can be communicated before representation is established. The intake script needs sign-off before deployment. Not a blocker, but a required step.

You are already booked out. If capacity is the constraint, capturing leads faster does not move the needle. Fix capacity first.

The matter type is highly variable. Practices where the first five minutes of every call are legal analysis — not fact collection — are harder to systematize. If the first call requires substantive judgment before you know what to ask, a call answering layer helps less.

The deployment I would build for a small firm#

Automated legal intake, in practice, is a working setup an owner can actually run: the AI answers the call, separates new leads from everything else, collects your practice-area intake fields, writes the structured note to your CRM, and routes or escalates by your rules — all before the caller hangs up. Here is the build for a solo or 2–5 attorney practice:

1. AI answers all new inbound calls and separates new leads from existing clients, courts, vendors, and other callers.
2. For new leads: collects practice-area-specific intake fields, captures opposing party names for conflict review, flags any deadline or urgency language the caller mentions.
3. Writes a formatted note directly to the firm’s CRM (Clio, Lawmatics, or MyCase) before the call ends.
4. Routes qualified leads to a consult booking or a human callback queue based on firm-defined rules.
5. Immediately alerts a human for any urgent escalation — arrest, imminent court date, safety issue.
6. Never gives legal advice.

That is a working AI-plus-CRM intake stack. It handles the administrative layer so the attorney and paralegal see complete, organized leads instead of voicemails that need a callback to fill in the basics.

If you want to see what this looks like built for your firm specifically, the free workflow audit is where to start. I will review your current intake questions, your CRM setup, and your call volume, then map exactly what to automate and what to leave with a human.

For more on the full attorney receptionist deployment, see Legal Intake AI Receptionist.

---

Sources referenced:

• Clio official pricing (Starter/Core/Signature/Elite; Grow included on Elite and Elite Work, add-on on Core and Signature; plans from $49/user/month, verified August 2026)
• MyCase help: How much does MyCase cost? ($50–$130/user/mo annual; $60–$150 monthly, dated June 1, 2026)
• Lawmatics pricing (custom plans, 3-user minimum, Essential 500 contacts / 10 automations, verified August 2026)
• Smith.ai AI Receptionist pricing (free 25-call tier, Pro from $150/75 calls, Enterprise from $500/300 calls, verified August 2026)
• Abby Connect pricing (Starter $165/50 Abby Minutes; AI uses half a minute, verified August 2026)
• Ruby plans and pricing (live receptionist minute plans; $250–$1,725 card cited from August 2026 recaps of ruby.com)
• eNZeti — intake coordinator salary 2026
• Smith.ai — law firm intake specialist pay by city (June 1, 2026)
• ABA Formal Opinion 512 on generative AI