Brightflag vs Onit

Brightflag and Onit both run legal spend and matter management, both point AI at invoice review, and as of mid-2026 both sell you a vendor-sourcing module. The decision is not the feature grid. It is whether you are buying one product metered on your outside counsel spend or a platform metered by module — and whether you want AI that hands your reviewer a summary or AI that goes and argues with the law firm.

What changed, and when

Wolters Kluwer completed its €425 million cash acquisition of Brightflag in June 2025. Brightflag had booked roughly €22 million in 2024 revenue, up 36%, and reached €27 million ARR by April 2025, with 155 employees moving across. Wolters Kluwer was explicit about the segmentation: its existing ELM Solutions business — TyMetrix 360° and Passport — serves large corporations and their law firms, and Brightflag extends the portfolio into mid-size corporations. Brightflag still ships under its own brand, its own domain, and its own roadmap.

Onit stayed independent under K1 Investment Management, which put $200 million in during 2019, and spent the same period rebuilding on an AI-native foundation. Spend Agent reached general availability in Q4 2025. Unity ELM — matter and spend on that foundation, with Ask Unity for querying the data — launched on 9 March 2026. Unity RFP followed on 11 May 2026. Onit has committed publicly to shipping one new AI-native product per quarter. It also bought LawBase in December 2025, a 46-year-old government case-management product, pushing into state, local, and education buyers.

Any comparison written before mid-2025 casts Brightflag as the independent AI challenger and Onit as a workflow roll-up bolting AI on top. Neither framing survives 2026.

Where Brightflag wins

• The bill does not move when the team does. A fixed one-time implementation fee plus an annual subscription sized to your outside counsel spend. Brightflag charges nothing extra for additional users or vendors, support, ongoing training, storage, bandwidth, or custom reports. Every paralegal, finance partner, and business-unit approver gets a login without a procurement conversation.
• Time to live has a number attached. Forrester’s Total Economic Impact study puts implementation at 6 weeks for first-time e-billing buyers and 10 weeks for teams replacing an incumbent. Finance integration — Workday, Oracle, or Coupa, by email, SFTP batch file, or two-way API — lands inside that window.
• PDF invoices are not a second-class path. The PDF Engine validates PDF invoices against matter data at the same accuracy as LEDES, flagging mismatched bill-to entities and purchase order numbers. Firms that refuse to convert billing formats do not become the hole in your spend data.
• The model is old, in the useful sense. Five shipped use cases run on a classification model trained over a decade: work classification against an extended UTBMS framework by both phase and activity, invoice summaries emailed to the reviewer, PDF validation, spend forecasting that returns a predicted date of budget overrun, and Ask Brightflag for querying the data conversationally.

Where Onit wins

• Scope past spend. Matter management, e-billing, legal holds, contract automation (AXDRAFT), deal rooms (SecureDocs), rate benchmarking (Bodhala), and government case management (LawBase) sit under one vendor relationship. Brightflag covers matters, vendors, and invoices — that is the whole surface, by design.
• The agent acts on the firm, not just the reviewer. Spend Agent identifies invoice errors, discrepancies, and billing-guideline violations, then interacts directly with the provider to adjust the invoice with minimal human intervention. Brightflag’s AI flags and summarizes for a human to decide. That gap is the largest functional difference between these two products.
• A two-sided network it has started to use. Onit puts its network at 4,000 corporate legal teams and 10,000 law firms. Unity RFP runs the sourcing lifecycle inside that network — vendor selection, rate intelligence, and matter execution in the same system — and Onit is releasing it to customers at no additional cost, staged from 1 June to general availability across the base in Q3.
• Cadence you can plan against. One AI-native product per quarter is a stated commitment, and Q4 2025 through Q2 2026 delivered against it.

The pricing math that actually differs

Neither vendor publishes a rate. Both route you to a quote, so the useful comparison is the meter, not a number you can look up.

Brightflag meters annual outside counsel spend. Onit meters modules and scope, with SimpleLegal — its mid-market spend product — priced on a percentage of legal spend. The consequence is arithmetic: two departments with identical outside counsel spend pay Brightflag the same whether one has 6 lawyers and the other has 30, while on Onit the larger department costs more and the price steps again every time a module turns on.

For the return side, Forrester’s composite reports 4% saved through billing-compliance enforcement (about $160,000 a year), 1% through better discount application (about $40,000), 0.75% on inaccurate timekeeper rates (about $30,000), and $45,000 from productivity — roughly $275,000 annually. Each of those percentages implies a composite organization with about $4 million in annual outside counsel spend, so read the study as a return of about 7% of spend, not as a dollar figure you can lift. Against that, entry mid-market Brightflag deployments land in the $25,000-40,000 per year range on our own buyer triangulation; treat it as an estimate, because the real number is a function of your spend tier.

Below roughly $2 million in outside counsel spend, neither product pays back. The flags fire on too few invoices to cover the subscription.

Ask Brightflag for the spend bands and the breakpoints between them. Ask Onit for the module list, the per-module quote, and confirmation that the same list is priced at renewal. Without those, the two quotes are not comparable.

Watch-outs

• Wolters Kluwer now sells two overlapping legal spend products. ELM Solutions for enterprise and Brightflag for mid-market is the stated split, but no published roadmap says what happens where they meet. Guard: ask which product the development spend goes to, get Brightflag’s roadmap commitments in writing, and confirm which legal entity holds the contract and what governs renewal.
• Onit’s AI layer is under a year old at every point. Spend Agent went GA in Q4 2025, Unity ELM shipped in March 2026, Ask Unity was in beta at announcement, and Unity RFP is still rolling out. Guard: require a reference customer already live on Spend Agent at your invoice volume, and pilot on one practice area before standardizing.
• An agent that negotiates with your panel is a relationship decision, not just a config. Guard: define which adjustment types it may make unattended, set a dollar threshold above which a human signs off, and tell your firms before it starts writing to them.
• Roll-up scar tissue on the Onit side. The portfolio was assembled from SimpleLegal, BusyLamp, AXDRAFT, SecureDocs, Bodhala, Legal Files, and LawBase; Unity is the consolidation push, not a finished state. Guard: confirm which SKU you are buying and whether the modules you need share one data model today.
• Spend-based pricing re-prices as you grow. Guard: negotiate the tier breakpoints and a renewal cap in the first contract, and model three years of spend growth — one acquisition can move you a tier.

Verdict

• Pick Brightflag if invoice review is the job to be done, your outside counsel spend runs roughly $5 million to $100 million, you want to be live this quarter, and you need a bill that does not re-price when headcount grows.
• Pick Onit if you need matter, spend, holds, and contracts under one vendor; if you want an agent that negotiates adjustments with firms instead of routing every flag to a human; or if you are a state, local, or education buyer, where LawBase has no equivalent on the Brightflag side.
• The default, if you cannot decide: Brightflag. Predictable metering and a six-to-ten-week go-live beat platform breadth for most departments, and the asymmetry favors it — adding matter management or CLM later is cheap, while unwinding an enterprise ELM program is not.
• If neither fits: Thomson Reuters Legal Tracker has the largest installed base of firms already submitting invoices, and Mitratech TeamConnect goes deeper on configuration if you have the IT capacity to run it. Under about $2 million in outside counsel spend, your AP system and a spreadsheet are the honest answer.

The mistake to avoid is comparing these on the AI demo. Both demos review an invoice and find violations. Only one of them then emails the firm — and whether you want that is a question about your outside counsel relationships, not about the software.

For the wider category, see best AI tools for legal ops; for where either fits alongside contract tooling, see the mid-market legal-ops stack.