Cohere Seeks $500 Million in Funding to Compete in AI Race

AI start-up Cohere seeks $500mn in effort to catch up to OpenAI and Anthropic

Artificial intelligence start-up Cohere is seeking to raise more than $500mn in new funding, aiming to catch up in the hugely expensive race to build cutting-edge AI models alongside rivals such as OpenAI, Google and Anthropic. The Canadian start-up is targeting a valuation of above $5.5bn, according to two people with knowledge of the discussions, broadly similar to the price tag it received from its previous funding round last year. Another person with direct knowledge of the discussions said the figure could reach between $6bn and $6.5bn, but cautioned the talks are in the early stages. While such a deal would still place it among the most valuable start-ups in the field, the talks suggest Cohere has fallen behind US competitors that have soared in comparison. In April, OpenAI secured a $300bn valuation, up from the $157bn in 2024, while Anthropic's latest funding round tripled its valuation to $61.5bn in March. Cohere was founded by former Google researchers, including chief executive Aidan Gomez. He is a co-author of Google's seminal "Attention Is All You Need" research paper, which introduced the concept of the transformer, an AI architecture that underpins all large language models. Unlike competitors, Cohere has not launched a consumer-facing app, and has positioned itself as a more enterprise and privacy-focused challenger to OpenAI and Anthropic. Cohere has also pushed into creating "open" models that are freely available for developers to access and build on, such as its Aya family of multilingual models. However, the company is entering a crowded market with alternatives from Meta and start-ups Mistral and DeepSeek. Cohere's founders Gomez, Nick Frosst and Ivan Zhang are also fighting for lucrative company contracts with tech giants such as Google, Microsoft and Amazon, which sell their own AI models to enterprises. The company has also doubled its annual recurring revenue in the last four months, crossing $100mn last month, according to a person close to Cohere. "A lot of the consumer adoption happened right away," said the person. "Enterprise tends to be slower in adoption but stickier in terms of users. Companies aren't known to adopt tech early." Developing powerful AI models also requires huge amounts of money to train models and pay for computing power. Nearly three years after OpenAI's ChatGPT kick-started the AI boom, investors are keen to see a return on the billions of dollars invested in hot AI model makers. A new challenger for Cohere is a crop of buzzy AI start-ups building applications on top of AI models, such as coding company Anysphere behind the popular coding tool Cursor. These start-ups are attracting keen investor interest and generating hundreds of millions of dollars of revenue in a matter of months. Earlier this year, Anysphere was valued at $2.5bn. Cohere has also introduced North, a platform that allows businesses to build AI agents for office work, but it is only available to a limited amount of customers.

Report: Cohere is Trying To Raise Another $500M As AI Race Speeds Up

Cohere, a generative AI startup, is seeking to raise over $500 million at a valuation of more than $5.5 billion, according to Financial Times. One source pegged the company's potential valuation to reach as high as between $6 billion and $6.5 billion. Founded by ex-Google researchers, Toronto-based Cohere is just one of many companies developing innovative AI models. It competes with the likes of OpenAI, Google and Anthropic. However, it has not created a consumer-facing app. Instead, it builds custom AI models, which do things like write website copy and power chatbots for enterprises such as Dell, Notion and Oracle. As its platform is cloud-agnostic, Cohere can be deployed inside public clouds, virtual private clouds, a customer's own cloud, or onsite. Since its 2019 inception, Cohere has raised about $1.1 billion, according to Crunchbase data. Backers include Nvidia, Alumni Ventures, Index Ventures, Radical Ventures and SentinelOne. Its last funding round was in July 2024, when it raised a $500 million Series D. One of Cohere's co-founders is Aidan Gomez, who co-authored Google's "Attention is All You Need," a technical paper believed to have laid the foundation for many of today's most capable generative AI models. According to an unnamed FT source, Cohere has doubled its annual recurring revenue in the past four months, crossing the $100 million mark in May. The company also competes with tech giants such as Google, Microsoft and Amazon, which also sell AI models to enterprises. In recent times, artificial intelligence startups have been getting the largest share of venture funding. Over the past year, nearly half of U.S. venture funding went to AI-related enterprises, Crunchbase data shows. AI dominates globally as well. Per Crunchbase's global funding report, AI was the leading sector for venture funding in Q1 2025, with $59.6 billion invested. The first quarter also came in as the strongest for AI funding ever, with a staggering 53% of global funding going to the AI sector alone.