Cohere raises a $125 million Series B

Cohere, the Toronto-based natural-language-processing company co-founded by former Google Brain researcher Aidan Gomez, raised $125 million in a Series B round led by Tiger Global, with returning investors Radical Ventures, Index Ventures and Section 32. The round took the company’s total funding past $170 million, roughly five months after its Series A.

Cohere’s pitch was access rather than novelty: an API that let businesses deploy large language models — for search, classification and generation — without building or hosting the models themselves. The company reported its user base had grown more than 800% since the Series A, and said headcount had doubled over the eight months before the announcement. The new funding was earmarked for platform development and international expansion, including a new Palo Alto office.

The round is notable less for its size than its timing. In February 2022, ChatGPT did not yet exist and OpenAI’s GPT-3 was still the reference point for what a commercial language model could do; Cohere’s raise, alongside similar rounds for other API-first NLP start-ups that year, signalled that investors were already betting a market would form around large language models beyond OpenAI itself — nine months before the launch that made that bet look obviously correct.