Top AI Companies in New Zealand 2026 (And Who's Actually Hiring)

Where the NZ AI scene actually lives

New Zealand's AI story is concentrated in four cities, and the weight isn't evenly distributed. Auckland holds about 60% of the named companies you'd care about. Wellington is the public-sector and fintech anchor. Christchurch quietly punches above its weight on aerospace and deep tech. Dunedin has a thin but real EdTech and bioinformatics cluster around Otago.

If you're a buyer trying to source an AI vendor here, or a multinational scouting partnerships, the map matters. The Auckland CBD and the Wynyard Quarter house most of the venture-backed AI startups. Wellington's Te Aro and Cuba Street strip is where you'll find Xero, Sharesies, Hnry, and the AI tooling that's grown around them. Christchurch's airport precinct and Tannery district are where Rocket Lab alumni and Dawn Aerospace people end up.

Related reading: Hiring AI Talent in New Zealand in 2026: An Employer's Playbook · Building an Engineering Hub in New Zealand: The Auckland-Wellington Playbook · Corporate Training Vendors in New Zealand 2026: An L&D Buyer's Guide.

The 12 NZ-headquartered companies doing serious AI work

Twelve isn't a magic number. It's just where the cut-off lands once you remove the ones still doing slide-deck AI. Each of these has shipped a product, has paying customers, and is hiring in 2026.

• Soul Machines (Auckland): Digital People, synthetic humans with their own "Digital Brain". Counts P&G, Mercedes-Benz, World Health Organization among customers. Raised a NZ$70M Series B-2 in 2022 and has been quietly scaling enterprise deals since.
• Halter (Auckland): Virtual fencing collars for cattle, with on-device ML that learns each animal's behaviour. Series D-ish funding closed in 2024 at around NZ$85M led by Bessemer. Hiring senior ML engineers and embedded firmware people through 2026.
• Auror (Auckland): Retail crime intelligence. Sequoia-led their Series B in 2022. Their model fuses CCTV, transaction, and law-enforcement data. The kind of cross-source ML work that sounds boring until you see the loss-prevention numbers.
• Xero AI (Wellington): Xero now treats AI as a product surface, not an R&D project. The "Just Ask Xero" assistant rolled out broadly in 2025 and the team in Te Aro is the largest applied ML group in the country.
• Tracksuit (Auckland): Brand-tracking platform with NLP at its core. Customer roster includes Liquid Death, Allbirds, Linktree. Growing fast off a US revenue base while keeping the engineering org in Auckland.
• Plexure (Auckland): Personalisation engine behind McDonald's app offers in dozens of markets. The kind of ML you don't notice until you realise every offer you see is targeted.
• Kami (Auckland): EdTech document collaboration with generative AI features for teachers. Used in something like 40,000+ schools globally; the AI feedback layer started rolling out in 2024.
• Crimson Education (Auckland): University admissions consulting with an AI matching layer for student-mentor pairings. Quirky business model, but the AI work under the hood is real.
• Mint Innovation (Auckland): Bio-based metal recovery from e-waste with ML-optimised bioleaching parameters. Less obviously "AI" but the optimisation work is heavy.
• Black Pearl Mail (Wellington): Email signature intelligence. Uses NLP to surface signal in outbound corporate email. Small team, big enterprise customers.
• Dawn Aerospace (Christchurch / Delft): Reusable suborbital spaceplane with autonomous flight stacks. Hires across NZ and the Netherlands; the autonomy work happens in both sites.
• Volpara Health (Wellington): Breast cancer detection ML, FDA-cleared, sold into US health systems. Acquired by Lunit in 2024 but the engineering team stayed in Wellington.

Honourable mentions: Imagr (computer vision for grocery checkout), Imagen AI (still operating quietly out of Auckland after the Rocket Lab spin-off), Partly (auto parts taxonomy, AI-heavy backend), and Sharesies' applied ML team in Wellington which doesn't market itself as an "AI company" but does substantial recommendation work.

Multinationals with AI teams on the ground here

This list matters more than the startup list for most B2B buyers, because these are the partners you'll actually be procuring through. The NZ presence ranges from genuine engineering centres to thin sales offices wearing the same logo.

• Microsoft NZ: The Auckland data centre region went live late 2024. Customer-facing AI services and a growing Azure AI specialist bench. Co-sell motion with NZ ISVs is the strongest of the hyperscalers here.
• AWS: The Auckland Region launched in 2025. Solutions architects with Bedrock and SageMaker depth are based in Auckland and Wellington.
• Google Cloud NZ: Smaller footprint than the other two, but the Vertex AI specialist team picked up notably in 2025 around the Auckland office.
• Datacom (NZ-listed, Auckland-headquartered): The largest indigenous IT services firm. AI practice has roughly doubled headcount in 2024-2025 and they're the default partner for Crown agencies doing AI POCs.
• Accenture NZ: Aggressive AI-services push, mostly Auckland-based, with the global Generative AI practice feeding work in.
• Deloitte and EY NZ: Both have built genuine AI consulting teams (not just slideware), though Deloitte is the larger of the two by headcount.
• Quantium NZ: The Australian data analytics firm runs a small but capable Auckland office that does retail and FMCG ML work.
• NVIDIA Inception: Not a direct hire footprint, but several NZ startups are in the programme and you'll see NVIDIA solutions engineers in Auckland for partner events.

What NZ AI companies are paying in 2026

Rough bands, in NZD, for Auckland and Wellington. Christchurch and Dunedin shave 5–15% off these numbers. Add ~10% on top for Soul Machines, Xero, Tracksuit, and the multinationals; subtract 5–10% for early-stage startups that lean on equity.

The honest read on these numbers: senior IC pay in NZ is roughly 30–40% behind Sydney for the same role, and the gap is wider against San Francisco or London. Equity at the Halter / Tracksuit / Auror tier closes some of it. The lifestyle premium closes the rest for the right kind of hire.

The Callaghan Innovation factor

Most foreign buyers underestimate how much Callaghan Innovation shapes the AI funding map here. The R&D Tax Incentive (RDTI) gives 15% back on eligible R&D spend; for a 20-person AI startup that's six figures of effective cash a year. The Ārohia Innovation Trailblazer Grant and Callaghan's project funding sit on top of that.

What this means in practice: NZ AI startups can afford a longer R&D arc on less venture capital than their Australian or US peers. Halter and Soul Machines both used Callaghan grants during early product work. If you're partnering or acquiring, factor this into the cap-table math, because it changes the dilution picture.

Callaghan itself has gone through restructure conversations in 2025-2026 and the agency is being narrowed. The RDTI sits with Inland Revenue and isn't going anywhere. The grants programme is the part to watch.

Where to start if you're a buyer or partner

A short list, in the order I'd actually use it:

1. NZTech and the AI Forum NZ member directories — both maintained, both useful for first-pass shortlists.
2. The annual AI Activity in New Zealand report from the AI Forum (typically published Q2 each year, 2026 edition due around now). Best single document on the state of the sector.
3. Callaghan Innovation's customer list — public, gives you a heat map of who's doing serious R&D regardless of marketing presence.
4. The Hi-Tech Awards finalists from the last two years — the AI categories are a reliable filter against vapourware.
5. Local meetups: AI/ML Auckland, Wellington AI, and the Christchurch ML group all still meet in person. You'll meet more useful people at one Wednesday session than from a week of LinkedIn outreach.

What about the talent crunch?

The honest constraint in NZ AI in 2026 isn't capital, it's senior ML talent. There are maybe 200 people in the country who can architect a production LLM system end-to-end. Most are already at the companies on this list. The Green List immigration pathway helps but it's slow; if you're standing up a team here, plan on 6–9 months to fill a senior bench, not 3.

That's not a reason to skip New Zealand. It's a reason to start the conversation early, to be specific about what you're hiring for, and to expect to compete on lifestyle and equity rather than just gross pay. The smart multinationals figured this out in 2023-2024. The ones arriving now are paying the late-mover tax.

What's worth watching for the rest of 2026

Four things are on my list for the back half of the year. The AWS Auckland Region's first full-year operating data will land mid-year, and the partner ecosystem (Datacom, Plexure, Tracksuit, several smaller ISVs) will start to show what's possible when you have a hyperscaler in-country. The next AI Forum NZ activity report (typically Q2-Q3) will sharpen the picture on which sectors are buying and which are still in pilot mode.

Halter's Series E rumours have been circulating since Q1; if those close they'll reset the bar for what a "scale" NZ AI startup looks like. And keep an eye on the Te Kawa Mataaho public-sector AI usage policies, currently in draft. How those land will shape how Crown-owned entities (NZ Post, Kāinga Ora, Te Whatu Ora) buy AI services for the next 18 months. None of these are individually a make-or-break event. Taken together, they're the difference between an AI sector that doubles in size by 2028 and one that quietly plateaus.