Multi-billion-dollar compute contracts with OpenAI, Microsoft, and Meta
Kubernetes-native infrastructure built for large-scale ML workloads
Data annotation and reinforcement learning from human feedback for AI training
Network of roughly 50,000 expert contributors worldwide
Reached $1.2B in annualized revenue entirely bootstrapped, no funding until 2025
Customers include OpenAI, Google, Anthropic, Microsoft, and Meta
First external fundraise in July 2025 targeted a $15B+ valuation
Pros
Dedicated AI-cloud focus avoids competing with general hyperscaler workloads for GPU capacity
Contracts with OpenAI, Microsoft, and Meta signal real enterprise-scale reliability
Public listing adds a level of financial transparency competitors lack
Trusted by nearly every major frontier AI lab for training data quality
Bootstrapped growth to $1B+ ARR demonstrates exceptional capital efficiency
Expert-contributor network built specifically for quality over crowdsourced volume
Cons
GPU-hour pricing scales fast at the training volumes CoreWeave targets
Built for large compute workloads, not a fit for small-scale experimentation
Heavy customer concentration among a few hyperscaler-adjacent contracts carries its own risk
Quote-based pricing requires a sales conversation, no published self-serve rates
Built for large-scale AI training projects, not suited to small one-off tasks
Primarily serves frontier AI labs rather than general business use cases
Company
CoreWeave, founded 2017
Surge AI, founded 2020
Prices come from each vendor's pricing page, last checked 20260716. Check the vendor before you buy.
FAQ
Questions people ask
Is CoreWeave better than Surge AI?
In our reviews CoreWeave scores 4.3 and Surge AI scores 4.3. The better pick depends on the job. CoreWeave: GPU cloud that IPO'd in 2025 and now powers training runs for OpenAI, Microsoft, and Meta. Surge AI: Reached $1B+ ARR bootstrapped, with zero venture funding until mid-2025.