Cohere Funding History
Cohere develops the Command family of large language models and tools that let companies ground model responses in retrieved information.
Company Overview
Cohere Funding Timeline
Series E (Initial Close; Round Remains Open)
Initial close; targeting up to CAD 7 billion
Cohere announced an initial close without disclosing the amount. Its CAD 35.7 billion valuation was conditional on reaching the full fundraising target. At the full fundraising target, the terms would imply a CAD 35.7 billion valuation.
Led by Schwarz Group.
Source · prnewswire.com ↗Funding Round
US$500 million
Cohere said the funding would support global expansion and development of enterprise and sovereign AI products.
Led by Radical Ventures.
Source · cohere.com ↗Series D
US$500 million
The company said the financing contained no secondary share sales, according to BetaKit.
Led by PSP Investments.
Source · en.wikipedia.org ↗Series B
US$125 million
Cohere planned to accelerate development of its natural language platform and expand its international presence.
Led by Tiger Global.
Source · en.wikipedia.org ↗Series A
US$40 million
Cohere said the financing would help it make natural-language models available to businesses through an API.
Led by Index Ventures.
Source · en.wikipedia.org ↗Series C
US$270 million
Cohere raised the round to expand its enterprise generative AI platform, with NVIDIA, Oracle and Salesforce Ventures among the strategic investors.
Led by Inovia Capital. Participation included NVIDIA, Salesforce Ventures, Index Ventures.
Source · globenewswire.com ↗
Cohere Founding Team
Aidan Gomez, Nick Frosst, and Ivan Zhang founded Cohere with experience in language model research. Gomez and Frosst had worked at Google Brain, and Gomez and Zhang had worked together in the independent research group FOR.ai.
Cohere Company Development
Enterprise Retrieval Became Part of the Model Offering
Cohere paired its Command family with retrieval tooling so applications could ground answers in enterprise material. That combined offering addresses the work of finding relevant information as well as generating a response from it.
Cohere Competitive Landscape
The Retrieval Stack Around the Model
Cohere and AI21 both give enterprise teams a model and a way to bring source material into the response. Teams can compare Command and Jamba in a retrieval workflow using their own documents, rather than judging the models on a standalone chat task.
