agentic legal unicorns - august 2025
on this page
point-in-time summary of harvey and legora as of august 2025. compiled from primary and secondary sources by google gemini and claude. many other relevant links available in bookmarks.
comparison
metric | harvey | legora |
---|---|---|
founded | 2022 | 2023 |
total funding | $806m | $116m |
valuation | $5b (jun 2025) | $675m (may 2025) |
arr at last val | $75m | $5-10m |
arr multiple at last val | 50-67x | 68-135x |
clients | 337 | 250+ |
current revenue | ~$100m | rumored ~$40-$50m |
revenue per client | $223-297k | $20-40k |
employees¹ | 642 | 130 |
revenue per employee | $117-156k | $50-350k |
linkedin followers¹ | 62,000 | 26,000 |
job openings¹ | 72 | 37 |
currently hiring in ¹ | nyc, sf, syd, lon | sthlm, nyc, syd, den |
¹ linkedin data retrieved august 6, 2025. employees = “associated linkedin members”
key metrics
valuation multiples: both companies at 50-135x arr vs. typical saas at 5-10x, anthropic ~20x, openai ~30x
revenue per client:
- harvey: $223k average (apr 2025: $75m arr / 337 clients), $120k minimum reported
- legora: $20-40k range (may 2025: $5-10m arr / 250+ clients)
growth: harvey raised $600m in 4 months (feb-jun 2025). legora went from yc to $675m in 1 year.
harvey
founded 2022 in san francisco by winston weinberg (ex-o’melveny & myers) and gabriel pereyra (ex-deepmind, google brain, meta). they were roommates when pereyra showed weinberg gpt-3.
funding history
date | round | amount | valuation | lead investors | source |
---|---|---|---|---|---|
nov 2022 | seed | $5m | - | openai startup fund | businesswire |
apr 2023 | series a | $21m | $150m | sequoia | businesswire |
dec 2023 | series b | $80m | $715m | kleiner perkins, elad gil | harvey |
jul 2024 | series c | $100m | $1.5b | google ventures | businesswire |
feb 2025 | series d | $300m | $3b | sequoia | bloomberg |
jun 2025 | series e | $300m | $5b | kleiner perkins, coatue | harvey |
revenue milestones
date | metric | value | source |
---|---|---|---|
dec 2023 | revenue growth | 10x in 2023 | harvey |
end 2024 | arr growth | 4x in 2024 | harvey year in review |
end 2024 | clients | 235 in 42 countries | harvey year in review |
end 2024 | am law 100 penetration | 28% | harvey year in review |
feb 2025 | arr | >$50m | businesswire |
apr 2025 | arr | $75m | techcrunch |
apr 2025 | clients | 337 | techcrunch |
aug 2025 | arr | $100m | cnbc |
mid 2025 | am law 100 penetration | 42% | pitchbook analysis |
valuation-to-arr multiple at last raise: 67x ($5b / $75m).
products
- assistant: conversational ai for document analysis and drafting (harvey)
- vault: bulk document storage and analysis (harvey)
- knowledge: legal research tool (harvey)
- workflows: agentic task orchestration launched june 2025 (harvey)
initially built on openai models. now uses multiple models from openai, anthropic, and google (harvey blog).
partnerships
microsoft (dec 2024): integration with sharepoint, onedrive, microsoft 365 copilot, teams, word. runs on azure. (harvey blog)
lexisnexis (jun 2025): integrating lexisnexis protégé service and shepard’s citations. co-developing motion workflows. (harvey blog)
pwc (sep 2024): custom tax ai model. 91% preference rate among tax experts. covers 10+ jurisdictions. (harvey blog)
pricing
per mlq.ai: starts at $1,200 per seat annually with 100-seat minimum ($120k base). custom models can exceed $5 million. median seat count doubles within 12 months.
clients
early adopter: allen & overy (now a&o shearman) (harvey newsroom)
current clients include: pwc, kkr, orrick, lowenstein sandler, macfarlanes, charles russell speechlys, comcast, verizon (harvey newsroom)
legora
founded 2023 in stockholm as “leya”. rebranded to legora february 2025. ceo max junestrand. developed inside mannheimer swartling for 9 months before commercial launch. (business wire)
funding history
date | round | amount | valuation | lead investors | source |
---|---|---|---|---|---|
apr 2024 | pre-seed | $0.5m | - | y combinator | yc profile |
may 2024 | seed | $10.5m | - | benchmark | legora blog |
jul 2024 | series a | $25m | - | redpoint ventures | bloomberg law |
may 2025 | series b | $80m | $675m | iconiq, general catalyst | tech.eu |
growth timeline
date | event | details | source |
---|---|---|---|
may 2023 | founding | built with mannheimer swartling | business wire |
jan 2024 | yc | winter 2024 batch | yc profile |
feb 2025 | rebrand | leya → legora, 200+ firms | business wire |
mar 2025 | us launch | nyc office, goodwin partnership | legora blog |
may 2025 | series b | $80m at $675m valuation | tech.eu |
jun 2025 | workflows | launched agentic workflows | legora blog |
arr estimated at $5-10m as of may 2025 based on client count and typical contract values.
products
- tabular review: spreadsheet-like document analysis (legora)
- assistant: conversational document q&a (legora)
- word add-in: microsoft word integration (legora)
- research: searches imanage, sharepoint, external databases (legora)
- workflows: agentic task automation launched june 2025 (legora)
built on microsoft azure. uses azure openai (gpt-4o) and anthropic claude. claude scored 18% better on their legal benchmarks.
partnerships
- cleary gottlieb (may 2025): firm-wide rollout after pilot (business wire)
- dentons (jun 2025): europe-wide deployment (dentons)
- goodwin (mar 2025): us market entry partnership (legora blog)
other clients: bird & bird, taylor wessing, pérez-llorca, mannheimer swartling (legora blog)
market data
metric | harvey | legora |
---|---|---|
founded | 2022 | 2023 |
total funding | $806m | $116m |
valuation | $5b (jun 2025) | $675m (may 2025) |
arr | $75-100m | $5-10m |
arr multiple | 50-67x | 68-135x |
clients | 337 | 250+ |
valuation multiples context
both companies trade at 50-135x arr. for comparison:
- anthropic: ~20.5x
- openai: ~30x
- public saas companies: typically 5-10x
legal services market size: >$1 trillion globally (pitchbook)
convergence on agentic workflows
both companies launched “agentic workflows” in june 2025:
- harvey: workflow builder announcement
- legora: workflows launch
these features enable ai to plan and execute multi-step legal tasks rather than just respond to queries.
sources
primary: harvey newsroom, legora blog
secondary: bloomberg, techcrunch, cnbc, business wire, tech.eu, dealroom, pitchbook, sacra