An engineer sets it up. Your firm sets the AI budget.
Misolla is priced in three parts: the setup, done for you by a forward-deployed engineer; the platform; and the AI your people use, drawn from one monthly pool your firm sets and shares out. You never prepay AI that sits unused.
How Misolla pricing works. Three parts, each priced for what it is.
Setting Misolla up around a firm is engineering work, running it is a subscription, and AI is a cost that rises and falls with the work. So each is priced on its own terms.
The setup
A forward-deployed engineer configures Misolla around your firm, connects what the work needs and trains your people. A defined amount comes with the subscription; substantial bespoke work is scoped and quoted before it starts.
The platform
The firm's workspace: matters, research, agents and the workflows built for you, with hosting, updates and ongoing support from the people who set it up.
The AI your people use
One monthly pool the firm sets, with a limit for each person inside it. The firm is billed for what is used, up to the pool.
A forward-deployed engineer. Not a login and a help centre.
Most of the value in legal AI is in fitting it to how a firm already works. That is engineering, so an engineer does it, with your people, on your matters. Your lawyers never have to become workflow builders.
Sits with the lawyers, paralegals and clerks who do it, and follows how a matter moves between them.
Finds every system the work touches: case management, document storage, email, calendars and accounts.
Connects the data sources a workflow needs, scoped with you, so work stops being retyped from one system into another.
Sets the workflows up on your own templates, precedents and policies, not generic ones.
On their own matters, when the firm decides to roll a workflow out.
Keeps tuning what is built after rollout, and takes on the next workflow the firm asks for.
Where the return comes from. The work between paralegals and lawyers.
In most law firms, the biggest gains are rarely a better answer to a legal question. They are in the repetitive work that passes between paralegals and lawyers: intake, chronologies, evidence packages, correspondence, filings. That is where the engineer starts, and where the result is measured.
Start with work that repeats
One piece of work the firm does often enough that doing it better shows up in the month, with a result a lawyer can judge.
Agree what to measure
Before anything is set up: time on the work, turnaround, how often a lawyer sends it back. The firm decides the criteria.
Judge it on your matters
The workflow is evaluated on the firm's own matters, not a demonstration. How a pilot works.
One AI pool for the firm. A limit for each person.
AI use is never spread evenly. A paralegal working through a 600-page medical disclosure uses far more than a lawyer asking a few research questions. So AI is not sold as the same allowance for every seat. The firm sets one monthly pool, gives each person a limit inside it, and keeps a shared reserve for the heavy work.
How the pool works. Set by your firm, not by us.
The firm sets the pool
The firm chooses its monthly AI budget, and can change it at any time.
Each person gets a limit
The firm gives each person, or each role, a monthly limit inside the pool. What isn't handed out stays as a shared reserve.
Everyone's use is visible
The firm sees use by person, matter and workflow as it happens, and is told when anyone passes a threshold it chooses.
Heavy work isn't cut off
When someone reaches their limit partway through a task, the task finishes, drawing on the shared reserve, and the firm is told.
Billed on use
The firm pays for the AI it actually uses, up to the pool. Budget that goes unused isn't charged.
No surprise overages
When the whole pool is spent, the firm decides: raise it, or let AI work pause until next month. Nothing is charged above the pool unless the firm raises it.
How this differs. From seat-based AI subscriptions.
Many legal AI products are sold per seat, with the same AI allowance for every person and setup left to the firm. Misolla is built the other way round.
| Topic | Seat-based AI | Misolla |
|---|---|---|
| Getting started | Accounts are switched on, and your people work out how to use them. | An engineer learns how your firm works and sets the workflows up around it. |
| Your systems | The tool sits beside them, and work is copied across by hand. | The data sources a workflow needs are connected, scoped with you. |
| AI usage | A fixed allowance per seat, paid for whether it is used or not. | One monthly pool the firm sets, billed on what is actually used. |
| Heavy work | One person hits their cap halfway through a long review. | The task finishes from the firm's shared reserve, and the firm is told. |
| Who controls spend | The vendor's plan tiers. | Your firm: the pool, each person's limit, and the alerts. |
| Proof before rollout | A demonstration on sample documents. | A pilot on your own matters, judged against criteria agreed at the start. |
Questions
- How does Misolla pricing work?
- In three parts: the setup, done by a forward-deployed engineer; a platform subscription; and AI usage, drawn from a monthly pool the firm sets, with a limit for each person. A defined amount of onboarding comes with the subscription, and substantial bespoke work or new connections are scoped and quoted before they start.
- Does Misolla publish its prices?
- Not during the private beta. Terms are agreed with each firm once the first workflow and its scope are clear, which is usually when a pilot is scoped.
- What is a forward-deployed engineer?
- An engineer from Misolla who works with your firm directly. They learn how the work runs between your lawyers and staff, connect the systems a workflow needs, set the workflow up on your own precedents, train your people, and keep supporting it after rollout. Your lawyers never have to build anything.
- How does the AI budget work?
- The firm sets one monthly AI pool and gives each person, or each role, a limit inside it. What isn't handed out is a shared reserve. The firm sees use by person, matter and workflow, and is alerted at a threshold it chooses.
- What happens when someone reaches their AI limit?
- A task already running finishes, drawing on the firm's shared reserve, and the firm is told. Long, token-heavy work such as reviewing a large medical disclosure is not cut off halfway.
- What happens when the firm's pool runs out? Are there overage charges?
- The firm decides whether to raise the pool or let AI work pause until next month. Nothing is charged above the pool unless the firm raises it.
- Do we pay for AI we don't use?
- No. The firm is billed for the AI it actually uses, up to the pool it set. Budget left at the end of the month isn't charged.
- Can we try it before committing?
- Yes. A pilot proves one workflow on the firm's own matters, judged against criteria agreed at the start. The firm decides whether to roll it out or stop, and keeps its data either way.
Join the firms building on Misolla.See it on a matter like yours.
Private beta. Concierge onboarding. Migration from your existing tools. Co-designed workflows for your practice.
Book a callConcierge onboarding · limited seats