Sunday 1 November 2026International EditionRSS
The financial daily of the startup economy
AI

Cohere's C$7bn round puts Ottawa on the cap table and a price on sovereign AI

Canada will take up to C$1.6bn of equity in Cohere's Series E, led by Schwarz Group. The terms say more about who now funds frontier AI outside America than the headline valuation does.

· · 5 min read
Chart for: Cohere's C$7bn round puts Ottawa on the cap table and a price on sovereign AI
Illustration: Invest Times

Cohere, the Toronto developer of large language models for businesses and governments, announced on Thursday 29 October the initial close of a Series E round that it hopes will reach up to C$7bn, at a valuation of C$35.7bn. The round is led by the companies of Schwarz Group, the German owner of Lidl and Kaufland. Bell Ventures, Bullhound Capital and the Government of Canada join as new investors, and Inovia Capital, Nvidia and Prosperity7 Ventures return. Cohere calls it the largest private capital raise in Canadian history.

The detail that sets this deal apart is the third new name on that list. On the same day in Montréal, Prime Minister Mark Carney said Ottawa would make an equity investment of up to approximately C$1.6bn, paid in stages and tied to commitments on jobs, computing capacity and the company's domicile. A G7 government is now a direct shareholder in one of the handful of independent model makers outside the United States and China. For investors in the sector, that changes both the funding arithmetic and the exit map.

Reading the valuation

The C$35.7bn figure needs care. Bloomberg described the size as up to $5bn and the valuation as $20.5bn before the new money. Converted at the exchange rate those two figures imply, the US dollar pre-money valuation corresponds to roughly C$28.7bn, and adding the full C$7bn target gives exactly the C$35.7bn Cohere has published. In other words, the headline number is a post-money figure that assumes the round fills completely. Cohere has not said how much was committed at the initial close; its chief financial officer, Francois Chadwick, told Bloomberg only that the first closing covered most of it.

On the pre-money measure, the step up is steep but not extravagant by the standards of this year. In August 2025 Cohere raised $500m in a Series D led by Radical Ventures and Inovia at $6.8bn. The new pre-money price is therefore roughly three times that level, a multiple Bloomberg also cited.

There is a second reference point. When Cohere signed its definitive agreement to combine with Aleph Alpha, the German model developer, in September, the merged company was valued at roughly $20bn, with Cohere shareholders keeping about 90%. Schwarz committed €500m at that stage to lead the coming round. The Series E pre-money valuation is essentially that merger value carried forward. The round is less a fresh repricing than a confirmation, by outside money, of a number the main investor had already agreed.

What the money buys, and what it does not show

Cohere's growth case rests on figures it chooses to disclose. It reports annual recurring revenue up 165% year on year at the end of the third quarter of 2026 and a customer base that grew 20 times over twelve months to 300. Named clients include Bell Canada, Fujitsu, RBC, TD Bank, PwC and the German naval shipbuilder TKMS. What it does not report is the level of that revenue. Chadwick declined to give a figure, and he declined to comment on a possible stock market listing.

That omission matters for anyone trying to put a multiple on the deal. The last widely cited data point, from a report by The Information in mid-2025, had Cohere expecting annualised revenue above $200m by the end of that year. Applying 165% growth to a base of that order would put recurring revenue in the hundreds of millions of dollars, not the billions. Even on generous assumptions, investors are paying a forward multiple that only makes sense if Cohere becomes the default supplier of AI to regulated industries and governments that will not, or cannot, rely on American or Chinese model providers.

The use of proceeds fits that thesis: model and product development, AI systems built for regulated sectors, long-term computing capacity to train frontier-level enterprise models, and expansion across North America, Europe and Asia. Schwarz Digits, the group's technology arm, wants Cohere's models to run on its STACKIT cloud. The co-founders frame the strategy as an Airbus to the American incumbents' Boeing, arguing in a company essay that the frontier "is not reserved for a fixed set of countries or companies."

The state as shareholder

The government's terms are the most consequential part of the announcement. According to the Prime Minister's Office, the investment is made on a commercial basis alongside the private backers, released in stages, and accompanied by regular reporting. In return Cohere commits to create at least 1,000 Canadian technical jobs by 2030, to spend at least C$1.7bn on additional computing infrastructure in Canada by the same year, to keep its Canadian incorporation and headquarters, to fund research partnerships and to keep working with the Canadian AI Safety Institute. The stake is being negotiated through the Canada Development Investment Corporation, the Crown corporation that manages government holdings.

For the company, a patient state shareholder lowers the cost of capital for the most expensive line in its budget, compute. It also strengthens its pitch to public sector buyers at home and in Europe, where provenance is now part of procurement. For other shareholders, the trade is optionality. Staged money conditioned on headquarters and incorporation makes a sale to a large American or Chinese technology group far harder to contemplate, at least while the tranches are outstanding. The exit routes that remain are a public listing, a combination with another non-American group, or a sale on terms Ottawa would accept. Late-stage investors entering at C$35.7bn are, knowingly or not, underwriting an IPO outcome.

A wider pattern of public capital

Canada is not acting alone. Germany's interest arrives through Schwarz rather than the federal budget, but the logic is similar: retailers, telecom groups and governments that once bought software are becoming financiers of the models they intend to depend on. Nvidia's return as an investor follows a pattern visible across the industry, where the main supplier of chips also sits on the cap tables of its customers. Taken together, Cohere's backers look less like a venture syndicate and more like a coalition of strategic and sovereign interests, each with its own reason to want the company to survive.

That is a strength in a downturn and a constraint in a bidding war. Coalitions protect a company against running out of money; they rarely maximise the price at which it is sold.

Our view

Cohere has found something many AI companies outside the largest labs lack: a group of backers whose interest is strategic rather than purely financial. That makes its survival more likely than its valuation alone would suggest. It does not make the valuation easy to judge. The headline is post-money, the amount actually closed is undisclosed and revenue is reported only as a growth rate.

For founders, the lesson is that sovereignty has become a fundable proposition, but it comes with covenants. Government equity tied to headquarters, jobs and compute spending narrows the set of future buyers, and every later investor will price that in. For investors, the useful questions are concrete: how much of the C$7bn is genuinely committed, whether recurring revenue is disclosed before the next raise, how quickly the Aleph Alpha combination closes and is integrated, and whether other governments follow Ottawa with equity rather than grants. If they do, a new class of shareholder will sit beside venture funds in the AI market, and it will measure returns in more than money.

Sources

Filed under AI. Written by Invest Times Editorial under our editorial standards. Commentary and analysis for information only, not investment advice.

More in AI

All AI analysis