Cohere secures initial Series E funding, targeting up to C$7 billion

Cohere has secured an initial close of a Series E targeting up to C$7 billion, putting new backing behind its bid to build advanced AI outside the United States and China. The company announced the financing on October 29, 2026, but did not disclose how much money had closed.

Schwarz Group leads the round. Named participants include Bell Ventures, Bullhound Capital, Inovia Capital, Nvidia, Prosperity7 Ventures and the Government of Canada. Cohere describes the financing as the largest round ever raised by a private Canadian company, a record claim made without publishing the size of the initial close.

The case for another AI builder

In a post signed by co-founders Nicholas Frosst, Ivan Zhang and Aidan Gomez, Cohere frames the investment as a response to dependence on a small number of AI providers. Its argument is that institutions should be able to inspect and adapt their systems, run them in their own environments, and choose technology governed by laws they trust.

The founders argue that access to essential technology can change because of commercial decisions, export controls or geopolitical events. They present the ability to build AI—not simply rent access—as a capability countries should retain. The financing therefore comes with a political proposition as well as a business plan: more allied countries and companies should be capable of supplying advanced AI.

What we are building is a second option for the democratic world, and the United States benefits from it as much as Canada, Germany, and other allies do.

Cohere co-founders Nicholas Frosst, Ivan Zhang and Aidan Gomez

Three destinations for the financing

Cohere says the money will support models, an institutional AI platform and cyber defenses. Its proposed route to competing is not to build one largest model for every job. Instead, it describes a range of models and surrounding software designed for different tasks, costs and deployment needs.

• Models at different price points. Cohere plans to keep investing in frontier models—systems intended to compete with the most capable available—alongside smaller, more efficient models. Customers would choose according to performance, cost, response time and where the system must run.
• A platform customers control. Cohere describes North as the layer governing an institution’s data and workflows. It says North runs in customers’ own clouds or data centers, including environments isolated from external networks, and supports Cohere models alongside models from other providers.
• Cyber defenses. The company identifies cybersecurity as a field where the complete system matters, not just the underlying model. Its argument is that software for exploring possibilities, testing suspected flaws and validating findings can contribute to results alongside raw model capability.

Cohere also describes a feedback loop between specialized work and general model development. Tasks with answers a computer can check, such as whether a test passes or a software bug can be reproduced, provide signals for improving models. Customer work, it says, exposes failures that public benchmarks may miss and helps identify which capabilities deserve further investment.

The founders acknowledge that competing at the frontier requires capital, computing resources and research talent at a scale few organizations can assemble. Their answer is to pursue several areas together: stronger general models, specialized systems and customer-controlled deployment. Cohere says the Series E is designed to accelerate that work; those are investment plans, not capabilities newly delivered by the financing announcement.