Cohere Secures $500M to Outpace Generative AI Rivals

Cohere, a generative AI startup that former Google researchers co-founded, has secured $500 million in new funding from a variety of investors, including Fujitsu, AMD, and Cisco

According to Bloomberg, the valuation of Toronto-based Cohere at $5.5 billion resulted from the round, including Canadian pension investment manager PSP Investments and Canada’s export credit agency EDC. The startup’s valuation has more than doubled since last June when it secured $270 million from Inovia Capital and other investors. This increases Cohere’s total to $970 million.

Josh Gartner, the head of communications at Cohere, stated to TechCrunch that the financing positions the company for “accelerated growth.”

According to a statement from Gartner, the company is “continuing to significantly expand our technical teams to develop the next generations of enterprise AI that are data privacy-focused and accurate.” “Cohere is dedicated to advancing the AI industry beyond esoteric benchmarks to provide tangible benefits in the daily workflows of global businesses across regions and languages.”

In March, Reuters reported that Cohere was discussing with Nvidia and Salesforce Ventures to raise between $500 million and $1 billion for its next fundraising round. In an email to TechCrunch, Gartner verified that both Nvidia and Salesforce ultimately contributed.

In 2019, Aiden Gomez, Nick Frosst, and Ivan Zhang founded Cohere. Gomez had previously conducted research at FOR.ai, which served as a precursor to Cohere. Gomez is a co-author of the 2017 technical paper “Attention Is All You Need,” which established the groundwork for numerous of the most advanced generative AI models today, such as Stability AI’s Stable Diffusion and OpenAI’s GPT-4o.

Unlike OpenAI, Anthropic, Mistral, and numerous generative AI startup competitors, Cohere does not prioritize consumer engagement. Rather, the organization personalizes its AI models, which are responsible for summarizing documents, writing website copy, and powering chatbots for customers like Oracle, LivePerson, and Notion.

Cohere’s AI platform is cloud-agnostic, capable of being deployed in public clouds (e.g., Google Cloud, Amazon Web Services), a customer’s existing cloud, virtual private clouds, or on-site. The startup employs a hands-on approach, collaborating with customers to develop customized models derived from their proprietary data.

In addition, Cohere operates a nonprofit research center, Cohere for AI, and makes available open-source models, such as multilingual models, to analyze and comprehend text. The Command R+, the organization’s most recent flagship model, is intended to offer many of the same features as more costly models (such as the GPT-4o) at a reduced cost.

Cohere’s strategy has been successful, even though both OpenAI and Anthropic are increasing their enterprise sales efforts. According to Bloomberg, Cohere was generating $35 million in annual revenue with a customer base of hundreds of companies at the end of March. This figure represents a growth from approximately $13 million at the end of 2023.

Generative AI is expensive at Cohere’s scale, especially as the company strives to develop more advanced systems. The new tranche and Cohere’s ongoing partnership with Google Cloud, in which Google provides cloud infrastructure to train and operate Cohere’s models, will undoubtedly be beneficial.

Oracle is both a customer and an investor in Cohere, and the startup’s AI is integrated into numerous Oracle software products, such as Oracle NetSuite.

Bloomberg reports that Cohere intends to increase its 250-employee workforce by 50% this year.