Coherea startup dedicated to Artificial Intelligence rival to OpenAI, and which is gaining a lot of relevance in recent weeks, has raised $270 million in a funding round. This amount comes from the sum collected by various venture capital funds, as well as by various investors who have taken the step strategically. Among them, Nvidia, Oracle and Salesforce Ventures.
With this round, Cohere, which is based in Toronto (Canada), is valued between 2.1 and 2.2 billion dollars, as confirmed by the company. In total, with the amount collected now, the total money obtained by Cohere in successive rounds adds up to 445 million dollars.
For Cohere, which develops and trains Artificial Intelligence models for companies, getting the necessary funding to establish itself and grow is a priority. Of course, they do not want to do it by giving too much power or control over the company to a single company.
So, Martin Kon, President and COO of Coderehas stated that they do not wantgive a large part of the company to another company, and then have the obligation to cover their needs ». Instead, they want to “make it accessible to as many companies as possible, on whatever cloud platform they choose.”«. Because Cohere’s technology is not tied to any particular cloud, and can be deployed and used across all. It is, in fact, available from any cloud provider you like. That is precisely why it is more attractive to many technology companies.
Since large language models cost a lot of money to train, and there is a lot of demand for computing power, which grows as these models become more popular, Cohere is looking to raise capital from companies that want to deploy the technology they develop for its users. Later, with the financing they obtain from them, they will have what is necessary to cover their computing needs, in addition to being able to expand towards the development of new technologies and companies from other sectors.
Cohere, founded in 2019, Its CEO and one of its founders is Aidan Gomez, an AI specialist who before founding the company worked for Google. Among other things, the company is engaged in the development of so-called large language models. In other words, software trained with huge amounts of data from the Internet so that they are capable of analyzing and generating text, also customizing it for the users who have requested it. Companies can use these models to automate certain routine tasks of their day to day, such as making email summaries or facilitating the writing of texts for various purposes.
