Cohere is one of the quietest unicorns in the AI industry. It has no viral consumer product and no consumer market, but in 2025 it achieved over 50% quarter-over-quarter growth throughout the year, reaching $240 million in annual recurring revenue (ARR). The question for Cohere is no longer whether it can succeed, but when it will go public.
The Numbers
- Annualized Revenue: $240 million (2025, exceeding the $200 million target)
- Total Funding: Approximately $1.5 billion
- Latest Valuation: $7 billion (after a $100 million top-up round in September 2025)
- Quarter-over-Quarter Growth: Above 50% throughout the year
These figures are solid for an enterprise AI startup. Cohere has no consumer traffic to fall back on; its growth comes entirely from paying B2B customers.
What They Sell
Cohere's clients include large enterprises such as Dell, Notion, and Oracle. Its core product lines:
| Product | Positioning |
|---|---|
| Command A / Command A Vision | Flagship generative models (including vision) |
| Embed 4 | Vector retrieval embedding model |
| Rerank 3.5 | Search re-ranking |
| North | Enterprise Agentic AI platform |
| Transcribe | Open-source speech recognition (new) |
North is the most critical recent product, positioned as a "safety-first Agentic AI platform" that addresses enterprise demand for controllable and auditable AI workflows.
Transcribe is the latest expansion — an open-source ASR model entering the speech recognition field.
Cloud-Agnostic Deployment as a Key Differentiator
OpenAI and Anthropic sell APIs that essentially push enterprise data onto their clouds. Cohere's strategy is different: it supports public cloud, private cloud, and on-premises deployment.
This is a hard requirement for industries like finance, healthcare, and government that are sensitive about data sovereignty. It is also the logic behind deep partnerships with Dell and Oracle — their customers are exactly these enterprises.
A Strong Resume
Cohere founder Aidan Gomez is a co-author of the 2017 paper "Attention is All You Need" — the paper that introduced the Transformer architecture. This credential carries significant weight in the AI community.
In the second half of 2025, Cohere hired Joelle Pineau, former Meta AI research vice president, as its Chief AI Officer. Pineau's academic background and experience with large-scale training at Meta should directly benefit Cohere's plans to build larger models.
CFO François Chadwick joined from Shield AI and has IPO experience. This appointment itself is an IPO signal.
IPO Outlook
CEO Aidan Gomez said in October last year that he "hopes to enter the public market soon." No timeline was given, but the CFO is in place, revenue is in place, and the valuation is not excessive — $7 billion reflects an enterprise SaaS logic, not a consumer AI bubble.
The current question is when the AI IPO window will open. OpenAI's IPO is set for Q4. If it goes smoothly, many will follow. Cohere could be among the first with a strong case — it has real B2B revenue and does not depend on consumer market sentiment swings.
The enterprise AI path has brought Cohere to this scale without a consumer product. It is a distinctly different route.
Sources: Enterprise AI startup Cohere tops revenue target as momentum builds to IPO (CNBC); CocoLoop, Enterprise GenAI Startup Cohere Confirms $500M Raise At $6.8B Valuation (Crunchbase News); Cohere Expands into AI Speech Recognition with Launch of Open-Source Transcribe Model (The AI Insider)