The “up to $65 billion” figure was Meta’s January 2025 forecast for its total capital expenditures that year—not a separate budget devoted only to artificial intelligence. Meta said the spending would support both generative AI and its core business, and that most of the 2025 capex would continue to go to the core business.
What did Meta mean by “up to $65 billion”?
In its January 29, 2025 earnings release, Meta forecast $60 billion to $65 billion in 2025 capital expenditures, including principal payments on finance leases. Capex is money spent on long-lived assets and infrastructure, rather than a standalone measure of all costs related to AI.
The range was a forecast for the full year, not an amount already spent. Meta said the planned increase would support its generative AI work as well as its core business. The company’s CFO Outlook Commentary stated: “The majority of our capital expenditures in 2025 will continue to be directed to our core business.”
What was the money expected to fund?
Meta’s January outlook identified infrastructure as the largest expected driver of expense growth, with compensation for technical talent the second largest. The capital-spending forecast therefore should not be read as a precise tally of AI costs: it covered the company’s broader infrastructure investment, and Meta did not publish a separate AI-only capex figure in that announcement.
#1 Best Overall
- Professional AI & Creator Workstation: AMD Radeon AI PRO R9700 GPU with 32GB GDDR6 is engineered for AI development, professional content creation, and compute-intensive workloads.
- Massive 32GB Memory Capacity: 32GB of GDDR6 memory on a 256-bit bus provides ample bandwidth for large AI models, 8K video editing, and complex 3D rendering.
- Advanced RDNA 4 with AI Accelerators: 64 Compute Units with 3rd Gen Ray Tracing and dedicated 2nd Gen AI Accelerators for groundbreaking AI performance and visual computing.
- Professional Blower Cooling: Efficient single blower design exhausts heat directly out of the chassis, ideal for multi-GPU workstation and server configurations.
- Enterprise-Grade Thermal Solution: Vapor chamber heatsink with industrial Honeywell PTM7950 thermal interface material ensures reliable cooling under sustained professional loads.
AI infrastructure can require substantial physical investment. Meta’s infrastructure overview describes a stack that includes its own MTIA accelerator chips aimed at inference, GPUs, CPUs, AI-optimized data centers, networking and the Research SuperCluster used for AI research. Meta says it designs, builds and operates infrastructure to tailor it to its workloads. That overview describes its approach; it does not identify which suppliers provided equipment associated with the $65 billion forecast.
At the time, Bloomberg reported that Mark Zuckerberg expected about one gigawatt of computing power to be online in 2025 and more than 1.3 million GPUs by year-end. Those were announced expectations, not independently verified year-end results. Bloomberg also reported Zuckerberg’s description of a planned data center as “so large that it would cover a significant part of Manhattan.”
Rank #2
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
How did Meta’s 2025 forecast change, and what did it spend?
| Date | 2025 capex figure | What it represents |
|---|---|---|
| January 29, 2025 | $60–65 billion | Initial full-year forecast, including principal payments on finance leases. |
| April 30, 2025 | $64–72 billion | Revised full-year forecast. Meta cited additional AI data-center investment and higher infrastructure-hardware costs; it said most capex would still support the core business. |
| January 28, 2026 | $72.22 billion | Reported full-year 2025 capex, including principal payments on finance leases. |
So Meta ultimately reported more 2025 capex than the original “up to $65 billion” forecast. The reported $72.22 billion is the full-year total under the company’s capex definition, not a disclosed amount spent exclusively on AI.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What is Meta’s latest capex outlook?
The 2025 figure is historical, not Meta’s current forecast. In its July 29, 2026 second-quarter release, Meta reported $31.08 billion in Q2 2026 capex and projected $130 billion to $145 billion in full-year 2026 capex. The company said it had narrowed its prior range of $125 billion to $145 billion. It attributed the higher lower end mainly to higher component pricing and, to a lesser extent, additional data-center costs for future capacity.
Rank #3
- Built for Running LLMs Locally: RDNA 4, 128 AI Accelerators, up to 1,531 TOPS (INT4) for fast inference and fine-tuning
- 32GB GDDR6 VRAM for Large AI Models: 256-bit, up to 640GB/s bandwidth, run large language and multi-modal AI models without offloading
- Multi-GPU Scaling for Local AI Clusters: PCIe 5.0 and 2-slot design support dense multi-GPU builds for local AI training and inference clusters
- Diecast Shroud and Backplate: Wave-pattern design cuts memory temperature by up to 16%, keeping clocks steady during long AI training runs
- Phase-Change GPU Thermal Pad: Delivers superior thermal conductivity for consistent performance and longevity under heavy AI loads
In that release, Zuckerberg said AI was “accelerating our core business today, powering our next generation of products, and opening the door to entirely new enterprise opportunities.” That is a statement of Meta’s strategy; it does not establish that the 2025 spending produced a particular financial return.
Quick Recap
Rank #4
- 24GB GDDR7 ECC Memory: handles large AI, 3D and rendering files smoothly
- Powerful CUDA Compute - 8,960 CUDA cores for fast graphics and computing power
- AI & Ray Tracing Boost - Tensor of the 5th generation and RT cores of the 4th generation
- PCIe 5.0 x16 interface - fast data connection with modern systems
- 4 × DisplayPort 2.1 - Multi-monitor support for professional workflows
How to interpret the headline number
- It was company-wide capex guidance: the January 2025 $60–65 billion range covered full-year capital expenditures, including finance-lease principal payments.
- It was not an AI-only allocation: Meta said spending supported both generative AI and its core business, with most capex continuing to be directed to the core business.
- It was a forecast that changed: Meta raised the range in April and later reported $72.22 billion for 2025.
- It does not establish investment returns: the company releases describe guidance and reported spending, not a guaranteed payoff or a quantified return from AI.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
