Harvey’s latest round tells you most of what you need to know about where legal AI stands. On September 9, the company raised $550 million at a $15.5 billion valuation. In December it had been worth $8 billion. Few corners of enterprise software have repriced that fast, and almost none of them sell to lawyers, a customer base that once treated a move to cloud email as a multiyear risk review.
The adoption numbers explain the money. 8am’s 2026 Legal Industry Report found 69% of legal professionals now use generative AI for work, up from 31% a year earlier. ILTA’s 2026 technology survey of 508 firms puts the share using or exploring it at 94%.
Yet the same 8am data shows only about a third of firms have rolled out legal-specific AI firm-wide, which means plenty of lawyers are pasting client work into tools nobody at their firm has vetted. Cost and ROI also broke into ILTA’s list of top concerns for the first time this year, named by 41% of firms. Buyers are past curiosity and are now asking what they get for the invoice.
That question has different answers depending on which part of legal work you look at. Here are seven of the top AI solutions for law firms, grouped by the job they do.
The platforms
Harvey is the default at large firms. The company says 80% of the Am Law 100 use it, and customers were running more than 25,000 custom agents on it as of March. It covers drafting, due diligence, research and review, and increasingly sells embedded legal engineers who build workflows inside client firms.
Legora, founded in Stockholm in 2023, is Harvey’s closest rival and has grown at a similar clip. It serves around 2,100 firms and legal teams across more than 80 countries, and Bloomberg reported last week that it is in talks to raise at roughly $8.5 billion pre-money. Its pitch leans on collaboration, including a client-facing Portal where firms publish their own workflows for clients to run themselves.
The incumbents with the library
Thomson Reuters’ CoCounsel Legal shipped a new version on August 20 that ties agentic workflows to Westlaw and Practical Law. That content is the moat, built from decades of editorially maintained case law that no startup can assemble on a Series B timeline.
Clio made the same bet from the other direction, buying vLex for $1 billion last year. Its Vincent AI now sits next to the practice management software many small and midsize firms already run, and in April Clio added agents that take goal-level prompts, such as finding everything that could sink a deal before signing.
The specialists
Spellbook lives inside Microsoft Word and sticks to contracts: drafting clauses, flagging aggressive terms, redlining against a firm’s playbook. The narrow focus has paid off; the company projects $100 million in ARR this year.
Altorney’s MARC goes after e-discovery, specifically the first-pass review that eats the bulk of a litigation budget. It classifies documents for responsiveness, privilege and personal data before they reach a hosting platform like Relativity, so firms pay to host far less. Founders Shimmy and Rachi Messing named it after their late father, Marc, an attorney.
On a Fortune 500 matter involving more than 200,000 documents, the company reports a 62% cut in review costs.
The front door
Upfirst.ai handles something the big platforms ignore: the phone. For a solo lawyer in court all morning, a missed call is frequently a missed client. Upfirst’s AI receptionist answers in more than 35 languages, runs intake scripts by case type, collects the opposing party’s name for a conflict check, and books consultations.
It was built by an attorney who got tired of paying someone to take messages, bills per call rather than per minute, and starts at $24.95 a month.
Disclosure: This article mentions clients of an Espacio portfolio company.
