Meta has announced a major expansion of its partnership with CoreWeave, committing approximately $21 billion to secure AI computing capacity over the long term.
The agreement, which extends through late 2032, is part of Meta’s broader strategy to scale its artificial intelligence infrastructure and meet rising demand for compute power.
Long-Term AI Infrastructure Deal
The partnership builds on a 2023 master services agreement and is structured as a capacity purchase deal, not an acquisition or equity investment.
Under this arrangement:
- Meta will pay for access to high-performance computing resources
- CoreWeave will supply AI infrastructure and cloud capacity
This reflects a growing trend where large tech companies rely on specialised AI cloud providers, often referred to as “neocloud” platforms.
Rising Demand for AI Compute Power
AI systems require massive computing resources for:
- Model training
- Inference (real-time response generation)
CoreWeave focuses on delivering infrastructure tailored for these workloads, offering both:
- Long-term contracted capacity
- On-demand usage models
Strategic Importance for CoreWeave
The Meta deal is significant for CoreWeave’s business diversification.
According to its 2025 filings:
- Revenue reached $5.1 billion
- Around 67% came from Microsoft
Adding Meta as a major client reduces reliance on a single customer and strengthens its position in the AI infrastructure market.
Expanding Ecosystem of Partnerships
CoreWeave has built a strong network of partnerships in the AI ecosystem, including:
- Multi-billion-dollar agreements with OpenAI
- Collaborations with companies like Perplexity AI and Poolside AI
- Close ties with NVIDIA, which has also made a significant equity investment
Meta’s Multi-Pronged AI Strategy
The deal aligns with Meta’s broader approach to AI infrastructure, which includes:
- Building in-house capabilities through initiatives like Meta Compute
- Developing custom AI chips
- Partnering with chipmakers such as AMD, NVIDIA, and Arm Holdings
- Securing external AI capacity through long-term contracts
Meta has also reportedly signed another AI infrastructure agreement with Nebius worth up to $27 billion.
Addressing Energy Challenges
Running large-scale AI systems requires significant energy resources. To support its infrastructure expansion, Meta has partnered with energy providers such as:
- Constellation Energy
- TerraPower
This ensures a stable power supply for its growing data centre network.
Focus on Inference Workloads
The CoreWeave deal is particularly important for AI inference workloads, which are becoming critical as AI systems move from development to real-world applications.
Serving millions of real-time queries often requires:
- Continuous computing capacity
- Scalable infrastructure
- High-performance cloud systems
External providers like CoreWeave help bridge capacity gaps as Meta continues to scale its own data centres.
Meta’s Latest AI Developments
Meta recently introduced Muse Spark, part of a new generation of large language models developed by its AI division.
This highlights the company’s aggressive push into AI, supported by both internal innovation and external infrastructure partnerships.
Industry Outlook
The deal reflects a broader shift in the AI ecosystem:
- Infrastructure providers like CoreWeave are becoming critical enablers
- Tech giants like Meta are emerging as large-scale buyers of compute capacity
- AI development is increasingly dependent on specialised cloud infrastructure
As demand for AI continues to grow, partnerships like this will play a key role in shaping the future of global AI infrastructure.
Ruchi Kumar is the associate editor at Entrepreneur News Network and TVW News India, where she leads editorial strategy, brand storytelling, and startup ecosystem coverage. With a strong focus on innovation, business, and marketing insights, he curates impactful narratives that spotlight India’s evolving entrepreneurial landscape. She has written extensively on fintech, AI and emerging startups.