Story of Cohere: offers NLP models that enhance communication between humans and machines
Story of Cohere:
In 2017, a research paper titled “Attention Is All You
Need” introduced a powerful new machine learning framework called the
Transformer. One of its co-authors was a young researcher named Aidan Gomez.
Inspired by the possibilities of this architecture, Gomez began to think about
how large language models could transform communication between humans and
machines.
Meeting of Minds
In 2019, Gomez teamed up with Ivan Zhang, a fellow researcher he had worked
with at FOR.ai, and Nick Frosst, who had been his colleague at Google Brain.
The three shared a vision: to build a company that would take cutting-edge
research in natural language processing and make it practical for developers
and businesses. Together, they founded Cohere, setting up headquarters
in Toronto, Canada.
Early Mission
From the start, Cohere wanted to make language AI accessible. Instead of
requiring companies to build their own expensive machine learning
infrastructure, Cohere offered ready-to-use NLP models through APIs. The goal
was to help developers integrate language tools like search, summarization, and
classification directly into their applications without needing deep ML
expertise.
Growing Team and Vision
Cohere’s culture was shaped by its founders’ backgrounds in research and
product development. They recruited engineers and scientists passionate about
exploring the potential of language AI. The team believed in combining research
with practical products to create tools that were not only powerful but also
useful across industries like finance, healthcare, and manufacturing.
Backing from Investors
Cohere quickly attracted the attention of prominent investors. Its backers
included Radical Ventures, Index Ventures, Section 32, Tiger Global, Salesforce
Ventures, Inovia, Oracle, Nvidia, SAP, AMD, and Cisco Investments, among
others. By 2024, the company had raised nearly $940 million across
multiple funding rounds, giving it the resources to scale globally.
Partnership with Google Cloud
In November 2021, Google Cloud partnered with Cohere to provide the
infrastructure needed for its AI models. Google’s TPUs became central to
Cohere’s operations, allowing the company to develop and deploy large-scale
language models efficiently while remaining cloud-agnostic for its customers.
Cohere for AI
In June 2022, Cohere launched Cohere For AI, a nonprofit research
initiative led by Sara Hooker, formerly at Google Brain. This lab was dedicated
to advancing open-source research in machine learning, enabling scientists
worldwide to contribute fundamental work that could shape the future of language
AI.
Expanding Capabilities
By December 2022, Cohere introduced a multilingual model that could understand
over 100 languages. Unlike earlier tools that were mostly limited to English,
this breakthrough allowed global businesses to use language AI for search and
discovery in their native languages.
Responsible AI Commitments
Cohere positioned itself as a responsible AI company. In September 2023, it
signed both the White House voluntary commitment on AI safety and
Canada’s AI code of conduct. These moves reflected the company’s
dedication to safe and transparent AI development.
Strategic Partnerships
Between 2023 and 2025, Cohere built partnerships with industry leaders. Oracle
integrated Cohere’s technology into its business applications, Salesforce used
its chat features, and McKinsey collaborated to help clients adopt generative
AI. Cohere also partnered with Fujitsu to build a Japanese LLM, RBC to launch a
banking AI platform called North, and LG to create a Korean version of North.
Story of Cohere: offers NLP models that enhance communication between humans and machines#startup#age#net worth#Cohere story#Entrepreneur#Cohere story
Expanding Worldwide
Cohere grew beyond Toronto, opening offices in San Francisco, London, and New
York City by 2024, and later expanding to Montreal in 2025. This global
presence positioned the company as a serious competitor to OpenAI, particularly
in serving enterprise clients with secure and customizable AI solutions.
Product Innovation
In March 2025, Cohere For AI introduced Aya Vision, a research-only
model capable of describing images, translating text, and summarizing
information. Around the same time, Cohere launched North, a secure AI
workspace platform tailored for regulated industries like banking and
healthcare.
Financial Growth
By March 2024, Cohere reported an annualized revenue run rate of $22 million,
which grew to $100 million by May 2025. Its valuation soared to $5.5
billion after securing $500 million in funding from Cisco, AMD, and
Fujitsu. The Canadian government also invested $240 million to support domestic
AI infrastructure.
Recognition and Awards
Cohere’s impact did not go unnoticed. It was featured on Forbes’ AI 50 list
four years in a row (2022–2025) and appeared on Fortune’s AI Innovators
list as well as CNBC’s Disruptor list. These accolades cemented its
reputation as one of the most influential AI companies in the world.
Challenges Ahead
Despite its success, Cohere also faced challenges. In February 2025, a group of
major publishers including The Atlantic and Condé Nast filed
lawsuits accusing Cohere of copyright infringement, claiming their articles had
been used for model training without permission. Balancing innovation with
legal and ethical concerns became part of the company’s journey.
Timeline of Cohere’s Story
- 2017
– Transformer architecture introduced; Aidan Gomez is a co-author.
- 2019
– Cohere founded in Toronto by Gomez, Zhang, and Frosst.
- Nov
2021 – Partnership with Google Cloud announced.
- Jun
2022 – Launch of Cohere For AI, a nonprofit research lab.
- Dec
2022 – Release of multilingual model covering 100+ languages.
- Sep
2023 – Signs White House AI safety commitment and Canada’s AI code of
conduct.
- 2023–2024
– Partnerships with Oracle, Salesforce, Fujitsu, McKinsey, LivePerson.
- Mar
2024 – Annual revenue run rate reaches $22M.
- Jul
2024 – Raises $500M funding at $5.5B valuation.
- Dec
2024 – Receives $240M public funding from Canadian government.
- 2024–2025
– Opens offices in New York City and Montreal.
- Jan
2025 – Launch of North for Banking with RBC and LG collaboration.
- Mar
2025 – Aya Vision research model released.
- May
2025 – Acquires Ottogrid in Vancouver; annual revenue hits $100M.
- Jun
2025 – Partners with Canadian & UK governments to expand AI in
public sector.
- 2025
– Faces copyright lawsuit from 14 publishing companies.
