Cohere is an enterprise-focused AI company building language, retrieval, coding, and deployment products, including the Cohere Command, North Mini Code, Embed, Rerank, North, and Model Vault families. Founded in 2019 by Aidan Gomez, a co-author of the original Transformer paper, and colleagues, it targets businesses and governments rather than consumers, emphasizing private deployments, data security, and sovereign AI. On April 24, 2026, Cohere and Aleph Alpha announced a planned combination, with Schwarz Group intending to back Cohere’s Series E with a EUR 500M structured financing commitment.
Key Facts
| Founded | 2019 |
| CEO | Aidan Gomez (Transformer paper co-author) |
| HQ | Toronto, Canada |
| Funding | About $1.54B raised before the 2026 Series E; Schwarz Group announced a EUR 500M structured financing commitment |
| Valuation | About $7B reported in 2025; current post-Series-E valuation should be verified from transaction documents |
| Models | Command, North Mini Code, Embed, Rerank, Transcribe |
| Focus | Enterprise, private deployment, sovereign AI |
| Revenue mix | Reported ~85% from private deployments |
What They Do
Cohere builds models tuned for enterprise needs: strong retrieval and embeddings for RAG, multilingual generation, coding-model infrastructure, and deployment options that keep data inside a customer’s environment. Rather than chase consumer chat, it sells to large organizations and governments where security, control, and sovereignty matter, with customers spanning finance, telecom, and enterprise software.
That positioning is the strategy. By concentrating on private and on-premise deployment, multilingual retrieval-strong models, open Apache 2.0 model releases, and European sovereign-AI relationships, Cohere differentiates from consumer-first labs and offers a credible option for regulated buyers who cannot send data to a public API.
Current Flagship Products
- Cohere, open weights, partner routes, and private deployment paths depending on model.
Strategic Position
Cohere’s moat is enterprise focus, retrieval strength, deployment flexibility, Apache 2.0 model-control signals, and a founder with deep model-research credibility. Its challenge is competing for enterprise spend against OpenAI, Anthropic, Google, and AWS-hosted options that also court regulated buyers. Its bet on private and sovereign deployment is a real differentiator but a narrower market than consumer AI.
For AIpedia readers, Cohere matters when the requirement is enterprise-grade, retrieval-heavy, privately-deployed AI rather than a consumer assistant.
Sources
- Cohere for AIpedia’s canonical product record.
- Cohere and Aleph Alpha join forces for the planned combination and Schwarz Group financing commitment.
- Cohere North Mini Code for the developer-model release.
- Cohere Command A+ for the open flagship model release.