What it is

Nuix is an Australian, ASX-listed (NXL) platform for processing and reviewing large, messy unstructured data: email stores, forensic images, encrypted and legacy files, chat exports. It sells two things that matter to a legal ops buyer. Nuix Neo is the processing and investigations platform, built on the engine that made Nuix Workstation a default for forensic and service-provider processing teams. Nuix Discover is the eDiscovery review application, sold as SaaS or on-premises. It competes with Relativity and Everlaw for review, and with Exterro FTK for forensics.

The company is healthier than it was. FY26 (year to 30 June 2026, reported 24 August) showed ACV of A$260.0M, up 13.9%, revenue of A$263.2M, and a statutory profit of A$16.4M against a A$9.2M loss the year before. Neo ACV rose 179% to A$78.5M across 135 customers. Nuix closed its up-to-€20M purchase of Linkurious, a Paris graph-analytics company, on 20 April 2026. John Ruthven became permanent CEO on 4 May 2026 after six months as interim. On 23 April 2026 the Federal Court dismissed ASIC’s disclosure case against Nuix and its directors; ASIC has appealed against the company only, and a shareholder class action began an eight-week hearing on 3 August 2026.

The question this page answers: is Nuix Discover a real alternative to RelativityOne or Everlaw, and is its AI usable? For processing-heavy investigations and on-premises review, yes. As a US litigation review platform picked on AI, not yet.

Why it shows up in Legal Ops stacks

Processing is the moat. Nuix reads 1,000+ file types, including corrupted, encrypted and legacy formats, at up to 1.5 TB an hour on its own figures, with deterministic output: the same inputs give the same results. Many service providers run Nuix processing and review in Relativity; Nuix is a Relativity ISV partner. If your problem is getting a forensic image or a 20-year-old mail archive into reviewable shape, Nuix is often already in the chain.
You choose where it runs. Discover ships as SaaS and on-premises, and Neo is built for local deployment. Government agencies, regulators and banks that cannot send case data to a US cloud pick Nuix for that reason, and Nuix called out strong government-sector performance in FY26.
The AI now covers review. On 4 September 2026 Nuix made Document Summaries (generated at ingestion), Semantic Search, Similar Documents, and AI Clustering and Visualizations generally available in Discover SaaS. On-premises follows next quarter. AI Chat, which answers plain-language questions about a matter with document citations and an audit trail, is in an early adopter program with GA set for December 2026. Nuix’s BYO AI lets you pick the model, including a local LLM.

Pricing reality

Nuix publishes no official price list (checked 2026-09-14). The public reference points:

Discover SaaS, single matter: $10 per GB per month with a $250 monthly minimum, per third-party listings.
Discover private cloud: no more than $4 per GB per month, falling toward $1 per GB at volume, per the same listings.
UK public sector: Nuix Discover SaaS is listed on G-Cloud 14 at £4.64 per GB per month.
Nuix Neo: custom enterprise licensing, usually multi-year. Multi-year deals were 35% of FY26 revenue.

The per-matter math: a 200 GB matter hosted for 12 months costs about $24,000 at the single-matter rate and about £11,000 at the G-Cloud rate. Processing speed matters as much as the hosting rate, because every day saved on processing comes off the matter timeline.

Best for

Investigations and eDiscovery leads at government agencies, regulators, financial institutions and service providers who process terabytes of mixed-format data and need it on-premises or in-country. Also for corporate legal teams that already run Nuix processing and want review on the same stack.

Not for you if you are a US litigation team choosing a review platform on AI maturity, if your matters depend on the large pool of Relativity-certified reviewers and service providers, or if you run a handful of small matters a year. Pick a self-service tool for those.

Versus the alternatives

RelativityOne: the market-share leader. Since early 2026, aiR for Review, aiR for Privilege and aiR for Case Strategy come inside standard RelativityOne per-GB pricing. Pick it for large US litigation and when outside counsel and service providers already work in Relativity.
Everlaw: the cloud-native growth challenger. Single-document AI Assistant actions have carried no extra charge since October 2025; batch AI actions still use credits. Pick it when review speed and usability decide the deal.
DISCO: since February 2026, one per-GB price on processed data with no ingest or user-license fees and unlimited Cecilia AI. Pick it when price predictability matters most.
Exterro FTK: pick it when forensics sits inside a wider legal hold and privacy program.

The direct answer: shortlist Nuix when processing volume or data sovereignty is the constraint, and run a live matter through both Nuix Discover and RelativityOne before you sign.

Watch-outs

Discover is the weaker half of the business. Discover is roughly an A$50M business, and SaaS revenue fell in FY26 after Nuix lost a large SaaS contract. Nuix has since set up a separate Discover team under its own EVP. Guard: ask for the Discover SaaS roadmap in writing, and cap the first term at one year with a renewal price ceiling.
The new AI is weeks old. Summaries and semantic search went GA on 4 September 2026; AI Chat is not GA until December. Guard: test the AI features on a closed matter where you already know the key documents, and write a defensibility check into the pilot.
Steep learning curve. G2 and Capterra reviewers of Discover and Workstation say the advanced tools take time to learn. Guard: budget vendor training for the first two matters and name one internal power user.
Legal overhang. ASIC’s appeal and the shareholder class action are still open. Guard: include data-return and escrow terms in the contract, and check the next ASX update before signing a multi-year deal.

Edit this page on GitHub