Canadian Deep Tech Startups Worth Watching in 2026
Canadian Deep Tech Startups Worth Watching in 2026
Founder Feast
September 4, 2026
Canada raised roughly $2.1B in deep tech venture funding in 2025, and about 60% of it landed in four verticals: quantum, synthetic biology, climate hardware, and semiconductors. That's a smaller pie than the US, but the concentration matters. When capital clusters, so do talent, exits, and follow-on rounds.
The interesting part isn't the headline number. It's that Canadian deep tech startups are no longer a Toronto-Waterloo story. Vancouver quietly became the country's quantum capital. Montreal owns AI hardware. Calgary is turning oil-and-gas engineers into carbon capture founders. Halifax is doing ocean tech that nobody in Silicon Valley knows exists yet.
Here's a grounded look at who's building, where the money is going, and which subsectors are actually producing venture-scale outcomes in 2026.
Quantum: Vancouver is the unlikely capital
D-Wave has been in Burnaby since 1999, but the last three years turned the Lower Mainland into a legitimate quantum cluster. Photonic Inc. raised $140M in 2023 from BCI and Microsoft for silicon spin qubits. 1QBit is quietly profitable doing quantum software for banks. Xanadu, technically Toronto-based, does most of its hardware work with UBC and SFU collaborators.
The talent pipeline is real. UBC's Stewart Blusson Quantum Matter Institute and TRIUMF pump out physics PhDs who would otherwise end up at IBM Yorktown or Google Santa Barbara. Instead, they're staying. Rent still matters, and a postdoc can afford Mount Pleasant in a way they can't afford Palo Alto.
The catch: quantum companies burn cash for a decade before revenue. Photonic has runway through 2027. D-Wave is public and trading at a fraction of its 2021 SPAC price. If you're a founder in this space, you need investors who understand 10-year timelines, which usually means strategics (Microsoft, Nvidia, DARPA-adjacent US agencies) rather than traditional Canadian VC. This is one reason many quantum founders end up doing a Delaware flip before Series B.
If you're building here, the BC startup ecosystem has more density in quantum and photonics than anywhere else in the country.
Synthetic biology and biotech: Toronto and Montreal split the pie
Deep Genomics raised $180M across its life. Ranomics, BenchSci, and Cyclica were all built in Toronto's MaRS orbit. The University of Toronto's Donnelly Centre is the single most productive biotech feeder in Canada, and the AbCellera exit in Vancouver (IPO'd at a $10B market cap, now trading lower but still a real company) proved you can build a billion-dollar biology company in Canada without moving to Boston.
Montreal has a different flavor. Valence Discovery (acquired by Recursion in 2023 for $47.5M) came out of Mila. Repare Therapeutics went public at $2B. The AI-plus-biology overlap is where Montreal punches above its weight, largely because Yoshua Bengio's shadow reaches into every drug discovery lab in the city.
Biotech in Canada has one persistent problem: capital efficiency looks great until you hit Phase II trials, and then you need $200M rounds that Canadian funds can't lead. Almost every biotech founder ends up raising from Boston, SF, or NYC. That's fine, but it means you need to know US investors early. If you haven't started building those relationships, this piece on Canadian founders and US investors is worth reading before your seed round closes.
Climate hardware: Calgary's second act
The most surprising cluster of 2025-2026 is Calgary. Ten years ago the city's talent pool was 90% oil and gas. Today, that same pool of chemical engineers, geophysicists, and drilling specialists is building carbon capture, geothermal, and hydrogen companies.
Carbon Upcycling raised $34M to turn CO2 into concrete additives. Eavor is doing closed-loop geothermal with $200M+ in strategic capital from BP and Chevron. Summit Nanotech pulled $50M for direct lithium extraction. These aren't consumer climate plays with marketing budgets. They're heavy-industry startups selling to utilities, cement companies, and mining majors.
Vancouver has its own climate hardware scene. Svante does carbon capture with $318M raised, including from Bill Gates' Breakthrough Energy. General Fusion has spent 20 years on magnetized target fusion and finally has a demonstration plant coming online. Moment Energy repurposes EV batteries for grid storage.
Climate hardware is capital-intensive, slow, and often the most rewarding for founders who understand physics and manufacturing. It's also the sector where Canadian federal funding (SIF, SDTC, NRC IRAP) actually moves the needle. A $10M SIF grant in climate hardware is the difference between building your pilot plant in Alberta or in Texas.
Semiconductors and hardware: quietly, Ottawa and Montreal
Nobody talks about Canadian semiconductors, but the country has a real chip design community concentrated in Ottawa (Ranovus, Ciena's spinouts, the old Nortel diaspora) and Montreal (Untether AI raised $125M for inference chips before hitting turbulence in 2025).
Tenstorrent, run by Jim Keller, is technically Toronto-based and now valued at $2.6B after its 2024 round. It's the closest thing Canada has to a hyperscale chip company, and it's hiring aggressively across Toronto and Austin. If Tenstorrent hits, it's the biggest Canadian hardware exit since RIM.
The subsector that nobody's noticed: photonics and optical networking. Vancouver-based EFFECT Photonics, Ottawa's Ranovus, and Toronto's POET Technologies are all building components that the AI datacenter buildout desperately needs. As GPU clusters scale past 100,000 chips, the bottleneck moves from compute to interconnect. Optical is where that gets solved.
For deeper context on how AI infrastructure and hardware intersect, the state of Canadian AI startups in 2026 covers the software side of the same story.
Where the exits are actually happening
Venture-scale outcomes in Canadian deep tech from 2020-2026, ranked by exit or last-round valuation:
- AbCellera (Vancouver, biotech): IPO'd at $10B, now $1.5B
- Tenstorrent (Toronto, semiconductors): $2.6B private
- Xanadu (Toronto, quantum): $1B private
- Photonic Inc. (Vancouver, quantum): estimated $500M+
- Repare Therapeutics (Montreal, biotech): $2B IPO, now smaller
- Eavor (Calgary, geothermal): $500M+ private
- Svante (Vancouver, carbon capture): $1B+ private
The pattern: biotech produces the biggest exits, semiconductors have the most upside per dollar invested, quantum is the longest-duration bet, and climate is where the strategic capital shows up early. If you're picking a subsector to build in, ask yourself which of those timelines matches your appetite.
Notice something else: outside of AbCellera, most of these companies eventually took majority US or international capital. Deep tech in Canada gets started here, but the follow-on rounds almost always require crossing the border. This is why the C-corp question matters so much for hardware and biotech founders specifically.
What Canadian deep tech founders are getting wrong
Three patterns come up repeatedly in conversations with technical founders across Vancouver, Toronto, and Kelowna.
First, they under-invest in commercial hiring. A quantum PhD founder will spend 18 months perfecting their tech stack and then hire their first salesperson six months before Series A. That's backwards. Your first commercial hire should happen the day you have a working prototype, not the day you need to close revenue.
Second, they treat government funding as the goal instead of the fuel. SIF, SDTC, and IRAP grants are useful, but they're not a business model. Every deep tech founder I talk to who's raised more than $50M in dilutive capital treats grants as a way to extend runway between VC rounds, not as their primary funding strategy.
Third, they wait too long to meet other technical founders. Deep tech is lonely. Your investors don't understand your tech, your friends don't understand your timelines, and your team is too small to have peers at your level. Founders who make it to Series B almost always have three or four other founder-friends they text at midnight when the reactor breaks. If you haven't built that bench, start now. Founder loneliness is real and it's worse in deep tech than anywhere else.
Common questions
Is Canadian deep tech actually competitive with the US? In specific subsectors, yes. Quantum (Vancouver), AI hardware (Toronto), and carbon capture (Calgary/Vancouver) are globally competitive. In general-purpose semiconductors and biotech Phase II+, no. Capital depth still matters and the US wins on Series B and later.
Do I need to move to be taken seriously? No, but you need to travel. Most successful Canadian deep tech founders spend 30-40% of their time in the Bay Area, Boston, or DC building investor and customer relationships. You can keep your team and IP in Canada while doing that.
What's the best Canadian city to start a deep tech company? It depends on the subsector. Quantum and climate: Vancouver. AI and semiconductors: Toronto or Montreal. Biotech: Toronto or Vancouver. Energy: Calgary. The best-province question has a longer answer, but for deep tech specifically, proximity to a research university with your specific specialty matters more than tax rates.
How much do I need to raise for a deep tech seed round in Canada? $3-8M is now standard for Canadian deep tech seeds in 2026. That's up from $1.5-3M five years ago. Hardware and biotech skew higher, software-heavy deep tech (quantum algorithms, bio-informatics) can still get away with $2M.
Where founders actually meet each other
Deep tech founders don't get much out of general startup events. The conversations that matter happen when you're sitting next to two other technical founders who are one round ahead or one round behind you, and neither of you has to explain what a qubit or a lipid nanoparticle is.
That's what we built Founder Feast for. Five founders, one restaurant, no pitching, Thursday evenings at 7pm in Vancouver, Toronto, and Kelowna. If you're building a Canadian deep tech startup and you want to spend an evening with four other people who actually understand what you're working on, apply here.
