Loading AI Valuations Dashboard...
AI startup and company valuations from OpenAI to emerging infrastructure players. Tracking $300M+ valued AI companies, funding rounds, and revenue multiples.
| Company | Valuation | ARR (est.) | Rev Multiple | Category |
|---|---|---|---|---|
| OpenAI | $300B | ~$5B | ~60x | Foundation models |
| xAI | $50B | ~$500M | ~100x | Foundation models |
| Anthropic | $61B | ~$1B | ~61x | Foundation models / safety |
| Databricks | $62B | ~$2.4B | ~26x | Data + AI platform |
| CoreWeave | $35B | ~$2B | ~18x | AI cloud infrastructure |
| Scale AI | $14B | ~$700M | ~20x | AI data / RLHF |
| Mistral AI | $6B | ~$50M | ~120x | Open-source models |
| Cohere | $5B | ~$100M | ~50x | Enterprise AI |
| Perplexity | $9B | ~$100M | ~90x | AI search |
| Harvey AI | $3B | ~$50M | ~60x | AI for legal |
| AI Category | Typical ARR Multiple | Range | Key Driver |
|---|---|---|---|
| Foundation model labs | 50–120x | Wide (revenue immature) | Strategic value, not revenue |
| AI infrastructure / cloud | 15–30x | More predictable | Contracted GPU revenue |
| AI data / RLHF | 15–25x | Maturing | Government contracts |
| Vertical AI (legal, medical, finance) | 20–50x | High variance | Net retention, defensibility |
| AI DevTools / APIs | 10–25x | Compressing | Usage growth, switching costs |
| AI-native SaaS (copilots) | 8–18x | Standard SaaS premium | Seat expansion, NRR |
Foundation model labs are valued primarily on perceived model quality and strategic optionality, not revenue. OpenAI's $300B valuation implies investors believe it will dominate the AI layer — not that $5B ARR justifies 60x today.
Proprietary training data and RLHF pipelines are increasingly the defensible asset. Companies like Scale AI are valued partly as data infrastructure — whoever controls high-quality human feedback data controls model quality.
AI companies embedded deeply into enterprise workflows (via APIs, agents, or co-pilots) command premium multiples because switching costs are high. Shallow AI wrappers on top of GPT-4 are seeing multiple compression as OpenAI commoditizes the layer below.
Most AI companies have mediocre gross margins today (40–60%) due to inference compute costs. Investors are betting on margin expansion as models get more efficient. Companies that can demonstrate improving margins with scale get premium valuations.
Recurring, contracted ARR commands a premium over project-based or usage-based revenue. OpenAI's ChatGPT subscription revenue and API ARR are treated differently in valuation models — sticky subscriptions are worth more than one-off API calls.
Microsoft's $13B investment in OpenAI, Google's $2B in Anthropic, and Amazon's $4B in Anthropic reflect strategic acquisition premium — big tech is essentially pre-buying AI infrastructure and optionality. This distorts private market valuations across the sector.
OpenAI's most recent valuation is approximately $300 billion, based on a funding round in early 2025. This makes it the most valuable private company in the world. With estimated annual recurring revenue of ~$5B, OpenAI trades at roughly 60x ARR — a premium that reflects expected model dominance and platform control, not current profitability. OpenAI is reportedly spending $7B+ per year on compute alone.
Anthropic was valued at approximately $61 billion in its most recent round in 2025. The company has raised over $12 billion total, with major investors including Google ($2B), Amazon ($4B+), and Spark Capital. Anthropic's ARR is estimated at ~$1B, implying a ~61x revenue multiple. Its focus on AI safety and enterprise reliability differentiates it from OpenAI's more consumer-forward positioning.
AI startup valuations are extremely elevated relative to current revenue, but the debate is whether they are bubbles depends on your time horizon. Foundation model labs (OpenAI, Anthropic) are valued on strategic optionality, not DCF math. Infrastructure plays (CoreWeave, Databricks) have real contracted revenue that partially justifies valuations. The most at-risk are thin AI wrapper companies — apps built on top of OpenAI APIs with no proprietary model or data advantage — which are seeing multiple compression as foundational AI capabilities commoditize.
AI companies in 2025 trade at significant premiums to traditional SaaS. Median public SaaS trades at 6–8x NTM revenue; top AI-native SaaS companies trade at 15–35x. Private foundation model labs trade at 50–120x ARR — multiples that have no precedent in traditional SaaS. The justification is market size potential: if AI infrastructure becomes the dominant compute layer, today's leaders could capture trillions in value.