---
title: "Meta Employee Count Statistics 2026: Headcount, Layoffs and AI Reallocation"
date: 2026-08-03
author: "Sofia Ramirez"
featured_image: "https://sqmagazine.co.uk/wp-content/uploads/2026/05/meta-employee-count-statistics.jpg"
categories:
  - name: "Technology"
    url: "/technology.md"
tags:
  - name: "Statistics"
    url: "/tag/statistics.md"
---

# Meta Employee Count Statistics 2026: Headcount, Layoffs and AI Reallocation

Meta Platforms employed 76,834 people at the end of December 2025, with approximately 50% based in the United States and the remainder across its global office network, per the FY2025 Form 10-K filing.

Headcount fell to approximately 74,700 by March 31, 2026, down approximately 3% from the 76,834 December 31, 2025, figure after a February 2026 reduction announced to redirect compensation budget toward Meta Superintelligence Labs and core AI infrastructure. The Layoff-to-AI-Hire Reallocation Arc from the 2022 peak of roughly 87,000 through four reduction waves and one [AI hiring](https://sqmagazine.co.uk/ai-job-creation-statistics/) surge is the workforce story this data tells. Reality Labs’ approximately $1.7 million operating-loss-per-employee ratio is the most extreme cost-per-head signal in big tech.

## Key Takeaways

- [Meta](https://sqmagazine.co.uk/meta-statistics/) employed **76,834** people as of December 31, 2025, broadly flat versus **74,067** at the end of 2024 after disciplined hiring in AI offset by reductions in other functions. Meta ended 2025 with 76,834 employees, broadly flat versus 74,067 at the end of 2024 after disciplined hiring in AI, infrastructure, and Reality Labs offset by reductions in other functions.
- Headcount fell to approximately **74,700** in Q1 2026 after the February 2026 announcement affecting approximately **7,500** roles. Meta announced it will cut 10% of its workforce in 2026, affecting approximately 7,500 employees with concentrations in recruiting, mid-management, and engineering functions not directly building AI products.
- RL spent approximately **$1.7 million** in operating loss per RL employee in 2025 against a **$20.6 billion** segment loss across approximately **12,000** staff, the highest per-head burn at any major US tech company. Reality Labs employs approximately 12,000 people; the implied operating loss per Reality Labs employee was approximately $1.7 million, the highest of any division at any major US tech company.
- Approximately **50%** of Meta’s workforce is US-based, with the rest spread across more than **40** cities globally. Meta has offices in more than 40 cities across the Americas, Europe, the Middle East and Africa, and Asia Pacific.
- Research and development expense totaled **$48.05 billion** in 2025, representing approximately **24%** of total revenue. Research and development expenses were $48.05 billion in 2025, an increase of 19% compared to $40.39 billion in 2024, and represented approximately 24% of total revenue in 2025.
- Meta Superintelligence Labs grew to approximately **1,300** researchers and engineers in its first year under Alexandr Wang. MSL now spans approximately 1,300 researchers, engineers, and product staff across Menlo Park, New York, London, and Paris offices.
- Cumulative reductions from November 2022 through March 2026 total approximately **29,100** roles across four waves, yet net headcount sits roughly at the 2023 baseline as AI hiring backfilled most cuts. The four announced layoff waves sum to approximately 29,100 roles: 11,000 (November 2022), 10,000 (March 2023), 7,500 (February 2026), and 600 (March 2026); each wave targeted recruiting, mid-management, and non-AI engineering while AI roles continued to grow.

## Editor’s Choice

- **76,834** total employees at the end of December 2025. As of December 31, 2025.
- **74,067** employees at the end of December 2024, an increase of **10%** compared to **67,317** at the end of 2023.
- Approximately **74,700** employees as of March 31, 2026, after the February 2026 reduction. As of March 31, 2026, headcount was approximately 74,700, down approximately 3% from 76,834 at December 31, 2025.
- Approximately **12,000** RL staff, per Reuters reporting, cross-referenced with Meta’s 10-K segment data. Reality Labs employs approximately 12,000 people across hardware engineering, software, and AI glasses operations.
- Approximately **1,300** employees in Meta Superintelligence Labs as of April 2026, per The Information’s reporting. Meta Superintelligence Labs now spans about 1,300 employees after Alexandr Wang’s first-year restructure.
- More than **40** cities house Meta offices, anchored by the global headquarters in Menlo Park. Meta has offices in more than 40 cities; the largest hubs include Menlo Park, New York, Seattle, Austin, Bellevue, Washington DC, London, Dublin, Paris, Tel Aviv, Singapore, Sydney, and Tokyo.
- **$200.97 billion** in FY2025 revenue. Full year 2025 revenue was $200.97 billion, up 22% year over year from $164.50 billion in 2024. Dividing FY2025 revenue ($200.97 billion) by the 10-K headcount (76,834) yields the implied revenue-per-employee figure.

## Recent Developments

- **April 30, 2026**: Meta reported Q1 2026 headcount of approximately **74,700**, down approximately **3%** from the **76,834** year-end 2025 figure. As of March 31, 2026, headcount was approximately 74,700, down approximately 3% from 76,834 at December 31, 2025.
- **April 15, 2026**: Meta Superintelligence Labs reached approximately **1,300** staff after Alexandr Wang’s first-year restructure, per The Information. Meta Superintelligence Labs now spans about 1,300 employees after Alexandr Wang’s first-year restructure, sources familiar with the org tell The Information.
- **April 8, 2026**: Meta announced Muse Spark, the first major proprietary AI model from MSL, replacing Llama as the inference engine of Meta AI. Muse Spark replaces Llama as the inference engine of Meta AI, which reaches 3 billion people across WhatsApp, Instagram, and Facebook.
- **March 12, 2026**: Meta cut roughly **600** additional recruiter roles in a second 2026 wave, per the Wall Street Journal reporting. Meta is cutting roughly 600 recruiting roles in a second wave of reductions announced internally on Tuesday.
- **February 28, 2026**: Mark Zuckerberg used Threads to defend the **10%** reduction as flattening management layers rather than contraction. Mark Zuckerberg used a series of Threads posts on Friday to defend Meta’s February 2026 workforce reduction, framing the cut as a flattening of management layers rather than a contraction.
- **February 4, 2026**: Meta announced the **10%** reduction of approximately **7,500** roles to redirect compensation toward Meta Superintelligence Labs and AI infrastructure. Meta Platforms will cut about 10% of its workforce; the cuts are expected to affect approximately 7,500 roles concentrated in recruiting, business operations, and engineering teams not directly building AI products.

## Meta Total Headcount Over Time

- **76,834** employees as of December 2025, broadly flat versus 2024 after AI-pivot hiring offset reductions elsewhere.
- **74,067** employees at the end of 2024, a **10%** increase from **67,317** at the end of 2023.
- Approximately **87,000** employees at the mid-2022 peak before the Year of Efficiency cuts began, per Bloomberg’s reconstruction. Across the two waves, Meta is reducing its workforce by approximately 25% from the peak of around 87,000 employees in mid-2022.
- The November 2022 wave removed **11,000** employees, primarily in recruiting, business partnerships, and product groups. The November 2022 layoffs removed 11,000 employees, primarily in recruiting, business partnerships, and product groups.
- The March 2023 Year of Efficiency wave eliminated another **10,000** roles, plus approximately **5,000** unfilled positions.
- Approximately **74,700** employees as of March 31, 2026, after the February 2026 cut, roughly equivalent to the 2023 baseline. As of March 31, 2026, headcount was approximately 74,700.

| Year | Headcount | % YoY Change | Key Event |
|---|---|---|---|
| 2022 (mid) | ~87,000 | n/a | Peak before efficiency cuts |
| 2022 (year-end) | ~76,000 | -13% | November 2022 wave (-11,000) |
| 2023 (year-end) | 67,317 | -11% | March 2023 Year of Efficiency wave |
| 2024 (year-end) | 74,067 | +10% | AI/infrastructure hiring resumed |
| 2025 (year-end) | 76,834 | +4% | Continued AI/RL hiring, modest non-AI cuts |
| Q1 2026 | ~74,700 | -3% (QoQ) | February 2026 10% cut |

*Source: Meta annual reports, Meta quarterly earnings releases, Bloomberg, Mark Zuckerberg public posts*

## Meta Employee Count Year over Year

- FY2024 ended at **74,067** employees, a **10%** increase compared to **67,317** at the end of 2023. As of December 31, 2024.
- FY2025 closed at **76,834**, a roughly **4%** increase over the FY2024 base. Comparing the 10-K year-end disclosures of 74,067 and 76,834 yields the implied annual growth rate.
- The increase in 2024 primarily reflected technical hiring in AI, infrastructure, monetization, and Reality Labs. The increase primarily reflects hiring in technical functions, including artificial intelligence, infrastructure, monetization, and Reality Labs, partially offset by reductions in certain business and operational teams.
- Q1 2026 saw the first sequential headcount decline since 2023, falling approximately **3%** to **74,700**. As of March 31, 2026, headcount was approximately 74,700, down approximately 3% from 76,834 at December 31, 2025.
- Severance and related restructuring charges tied to the February 2026 reduction are expected to total approximately **$1.5 billion** in H1 2026. Severance and related restructuring charges are expected to total approximately $1.5 billion in the first half of 2026.
- Capital expenditures rose to **$72.22 billion** in 2025 and are expected to be **$115 billion to $135 billion** for FY2026, the highest in Meta’s history and explicitly tied to AI infrastructure. Meta guided 2026 capital expenditure to $115 billion to $135 billion to support Meta Superintelligence Labs and core business AI infrastructure, up from 2025 capex of $72.22 billion.

| Period | Headcount | YoY/QoQ Change | Primary Driver |
|---|---|---|---|
| FY2023 | 67,317 | -11% YoY | Year of Efficiency completion |
| FY2024 | 74,067 | +10% YoY | AI / infra / RL hiring |
| FY2025 | 76,834 | +4% YoY | Continued AI hiring, disciplined elsewhere |
| Q1 2026 | ~74,700 | -3% QoQ | February 2026 10% cut |

*Sources: SEC EDGAR filings (Meta 10-K), Meta IR earnings releases*

## Family of Apps vs Reality Labs Headcount Split

- Family of Apps generated **$193.7 billion** in 2025 revenue, the bulk of Meta’s **$200.97 billion** total. Family of Apps revenue was $193.7 billion.
- RL generated **$2.95 billion** in 2025 revenue with an operating loss of **$20.6 billion**. Reality Labs revenue was $2.95 billion with an operating loss of $20.6 billion for the full year.
- RL employed approximately **12,000** people in 2025, per Reuters tracking of internal sources cross-referenced with the FY2025 10-K. Meta’s Reality Labs division employs approximately 12,000 people across hardware engineering, software, and AI glasses operations.
- Family of Apps consequently accounts for roughly **64,800** of the **76,834** total headcount, implying approximately **84%** of the workforce sits in the money-making segment. Subtracting the 12,000 RL headcount from the 76,834 10-K total leaves the implied Family of Apps headcount.
- RL revenue rose **48%** year over year from **$2.0 billion** in 2024. Reality Labs revenue was $2.95 billion in 2025, an increase of 48% compared to $2.0 billion in 2024.
- Family of Apps’ operating income rose to $87.1 billion in 2025 from $64.0 billion in 2024. Family of Apps’ operating income was $87.1 billion in 2025, up from $64.0 billion in 2024.

![Family Of Apps Vs Reality Labs Headcount](https://sqmagazine.co.uk/wp-content/uploads/2026/05/family-of-apps-vs-reality-labs-headcount.jpg "Family of Apps vs Reality Labs Headcount")

### What Percent of Meta Works on Reality Labs?

Comparing the 12,000 RL headcount estimate against the 76,834 10-K total yields the implied RL workforce share, with the balance sitting in Family of Apps and corporate functions.

## US vs International Employee Mix

- Approximately **50%** of Meta employees are based in the United States, with the rest spread across the global office network.
- The same **50%** split held at the close of 2024, indicating a stable geographic mix despite the headcount churn.
- Meta operates in more than **40** cities globally. Meta has offices in more than 40 cities across the Americas, Europe, the Middle East and Africa, and Asia Pacific.
- Major US hubs include Menlo Park (headquarters), New York, Seattle, Austin, Bellevue, and Washington DC.
- International engineering is concentrated in London, Tel Aviv, and Paris. Engineering and AI research are concentrated in Menlo Park, Seattle, Bellevue, New York, London, Tel Aviv, and Paris.
- Reality Labs engineering operates from Burlingame, Redmond, Sunnyvale, and Zurich. Reality Labs engineering operates from Burlingame, Redmond, Sunnyvale, and Zurich.

| Region | Key Cities | Primary Function |
|---|---|---|
| North America (US) | Menlo Park, NYC, Seattle, Austin, Bellevue, Washington DC | HQ, engineering, AI research, ops |
| Europe / EMEA | London, Dublin, Paris, Tel Aviv | Engineering, sales, AI research |
| Asia Pacific | Singapore, Sydney, Tokyo | Sales, partnerships |
| US (Reality Labs) | Burlingame, Redmond, Sunnyvale | Hardware engineering |
| Europe (RL) | Zurich | Hardware engineering |

*Sources: Meta Careers global office network, SEC EDGAR (Meta 10-K)*

## Engineering vs Non-Engineering Headcount Mix

- Meta tracks retention across major functions, including engineering, product, sales, and operations, per the FY2025 10-K.
- The 2024 hiring surge was concentrated in technical functions, including [AI](https://sqmagazine.co.uk/artificial-intelligence-statistics/), infrastructure, monetization, and Reality Labs. The increase primarily reflects hiring in technical functions, including artificial intelligence, infrastructure, monetization, and Reality Labs.
- R&D expense represented approximately **24%** of FY2025 revenue, driven by payroll for engineering, research science, and product employees, including MSL hires. Research and development expenses were $48.05 billion in 2025; R&D represented approximately 24% of total revenue in 2025, primarily driven by payroll and related expenses for additional engineering, research science, and product employees, including hires associated with Meta Superintelligence Labs.
- The 2026 cut targeted engineering teams not directly building AI products, leaving AI engineering largely insulated. The cuts are expected to affect approximately 7,500 roles concentrated in recruiting, business operations, and engineering teams, not directly building AI products.
- The Year of Efficiency 2023 cuts focused on middle management and engineering managers whose work overlapped with other functions. The cuts target middle management and product teams whose work overlapped with other functions, with engineering managers a particular focus.
- Mark Zuckerberg signaled in the Q4 2025 earnings call that AI will dramatically change how employees work in 2026, with teams flattening and adopting internal AI tooling. Mark Zuckerberg said AI will dramatically change the way employees work in 2026 as teams flatten and adopt internal AI tooling.

| Function | Engineering Share Signal | Source Type |
|---|---|---|
| AI / MSL | Hiring continues through 2026 | Meta IR, The Information |
| Family of Apps engineering | Stable; non-AI engineering trimmed Feb 2026 | Reuters, Meta IR |
| Reality Labs engineering | ~12,000 RL staff, largely engineering | Reuters, Meta 10-K |
| Recruiting / Talent Acquisition | Cut ~600 in March 2026, on top of Feb 2026 | WSJ |
| Mid-management | Heaviest 2023 hit; partially repeated in 2026 | Bloomberg, Reuters |

*Sources: SEC EDGAR (Meta 10-K Item 1), Reuters, Wall Street Journal, The Information, Bloomberg*

### How Many Engineers Does Meta Actually Have?

Meta does not publish an engineering-only headcount in its SEC filings. The 10-K names engineering as one of four “major functions” alongside product, sales, and operations, without a numeric split. External estimates from LinkedIn and trade reporting put engineering at roughly 35,000-45,000 staff during 2024-2025, but those figures are not officially confirmed.

## Research and Development Headcount Share

- R&D expense was **$48.05 billion** in 2025, a **19%** increase compared to **$40.39 billion** in 2024. Research and development expenses were $48.05 billion in 2025, an increase of 19% compared to $40.39 billion in 2024.
- R&D represented approximately **24%** of total revenue in 2025. R&D represented approximately 24% of total revenue in 2025.
- The R&D increase was primarily driven by payroll for additional engineering, research science, and product employees, including MSL hires. R&D was primarily driven by payroll and related expenses for additional engineering, research science, and product employees, including hires associated with Meta Superintelligence Labs.
- FY2025 revenue was **$200.97 billion**, up **22%** year over year from **$164.50 billion** in 2024. Full year 2025 revenue was $200.97 billion, up 22% year over year from $164.50 billion in 2024.
- Capital expenditures rose to **$72.22 billion** in 2025 and are expected to be between **$115 billion and $135 billion** for FY2026 to support MSL and core AI infrastructure. Meta guided 2026 capital expenditure to $115 billion to $135 billion to support Meta Superintelligence Labs and core business AI infrastructure, up from 2025 capex of $72.22 billion.
- Q1 2026 capital expenditures totaled **$29.4 billion**, primarily directed toward AI training capacity. Capital expenditures in Q1 2026 totaled $29.4 billion, primarily directed toward AI training capacity.

![R&D Spending vs Revenue Growth Trends](https://sqmagazine.co.uk/wp-content/uploads/2026/05/r-d-spending-vs-revenue-growth-trends.jpg "R&D Spending vs Revenue Growth Trends")

## AI and Generative AI Team Growth

- Meta Superintelligence Labs was formed in **June 2025** to combine FAIR, the GenAI organization that built Llama, and a new Superintelligence team focused on long-horizon research.
- Alexandr Wang, founder and former CEO of Scale AI, leads MSL as Chief AI Officer. Alexandr Wang, founder and former CEO of Scale AI, will lead MSL as Chief AI Officer.
- Roughly **600** MSL employees focus on foundation modeling and Muse Spark research. Roughly 600 employees focus on foundation modeling and Muse Spark research.
- Approximately **400** MSL employees work on applied AI for the Family of Apps, including the Meta AI assistant. Around 400 work on applied AI for the Family of Apps, including the Meta AI assistant.
- Alexandr Wang has brought in dozens of senior hires from [OpenAI](https://sqmagazine.co.uk/openai-statistics/), Anthropic, and Google DeepMind during the past nine months. Wang has consolidated previously scattered AI teams and brought in dozens of senior hires from OpenAI, Anthropic, and Google DeepMind during the past nine months.

> **By the numbers:** Meta Superintelligence Labs spans approximately **1,300** researchers and engineers as of April 2026, per The Information’s reporting on Alexandr Wang’s first-year restructure. Roughly **600** focus on foundation modeling, **400** on applied AI for the Family of Apps, and the remainder on long-horizon research and infrastructure support.

For prior coverage of Meta’s AI capabilities and adoption metrics, the live [Meta AI statistics](https://sqmagazine.co.uk/meta-ai-statistics/) page tracks the user and capex side of the same pivot.

![AI Workforce Distribution Across Key Initiatives](https://sqmagazine.co.uk/wp-content/uploads/2026/05/ai-workforce-distribution-across-key-initiatives.jpg "AI Workforce Distribution Across Key Initiatives")

### How Big Is the Llama Team at Meta?

The [GenAI](https://sqmagazine.co.uk/generative-ai-statistics/) organization that built Llama was folded into Meta Superintelligence Labs alongside FAIR and a new Superintelligence team. Llama persists as the prior model family, but the active team headcount question now maps to the foundation-modeling pillar of approximately 600 staff per The Information.

## Meta Superintelligence Labs Headcount

- MSL launched in June 2025 with Alexandr Wang as Chief AI Officer.
- MSL reached approximately **1,300** staff by April 2026 after a first-year restructure. Meta Superintelligence Labs now spans about 1,300 employees after Alexandr Wang’s first-year restructure.
- The lab is organized into three pillars: foundation modeling, applied AI for Family of Apps, and long-horizon research. The lab is organized into three pillars: foundation modeling, applied AI for the Family of Apps, and long-horizon superintelligence research.
- Muse Spark, MSL’s first major proprietary model, launched on April 8, 2026, and replaced Llama as the Meta AI inference engine.
- Meta AI reaches **3 billion** people across WhatsApp, Instagram, and Facebook, the install base MSL builds for. Muse Spark replaces Llama as the inference engine of Meta AI, which reaches 3 billion people across WhatsApp, Instagram, and Facebook.
- MSL hiring continues to outpace the rest of Meta even as the broader company sheds roles. MSL hiring continues to outpace the rest of Meta even as the broader company sheds roles.

For comparison context, the published [OpenAI employee count](https://sqmagazine.co.uk/how-many-people-work-at-openai/) tracks the most similar specialist-AI competitor by headcount and trajectory.

| Pillar | Approx. Staff | Lead Office Hubs |
|---|---|---|
| Foundation modeling | ~600 | Menlo Park, Paris |
| Applied AI (Family of Apps) | ~400 | Menlo Park, New York |
| Long-horizon research + infra | ~300 | London, New York |
| Total MSL | ~1,300 | Four-city network |

*Source: The Information, Meta Newsroom*

## Meta Layoffs Timeline

- The November 2022 wave removed **11,000** employees, primarily in recruiting, business partnerships, and product groups. The November 2022 layoffs removed 11,000 employees, primarily in recruiting, business partnerships, and product groups.
- The March 2023 Year of Efficiency wave eliminated **10,000** roles plus approximately **5,000** unfilled positions.
- The February 2026 wave affected approximately **7,500** roles concentrated in recruiting, business operations, and non-AI engineering. The cuts are expected to affect approximately 7,500 roles concentrated in recruiting, business operations, and engineering teams, not directly building AI products.
- The March 2026 second-wave cut roughly **600** additional recruiter roles. Meta is cutting roughly 600 recruiting roles in a second wave of reductions announced internally on Tuesday.
- Across the November 2022 and March 2023 waves, Meta reduced its workforce by approximately **25%** from the mid-2022 peak of around **87,000**. Across the two waves, Meta is reducing its workforce by approximately 25% from the peak of around 87,000 employees in mid-2022.
- Severance and related restructuring charges for the February 2026 cut are expected to total approximately **$1.5 billion** in H1 2026. Severance and related restructuring charges are expected to total approximately $1.5 billion in the first half of 2026.

> **Why it matters:** Across four reduction waves from November 2022 through March 2026, Meta has eliminated approximately **29,100** roles, roughly **33%** of the **87,000** peak workforce, yet net headcount has settled at approximately **74,700** as AI hiring backfilled most of the cuts. The waves consistently targeted recruiting, mid-management, and non-AI engineering.

| Date | Count | Function Focus | Severance / Notes |
|---|---|---|---|
| November 2022 | 11,000 | Recruiting, business partnerships, product | First mass layoff |
| March 2023 | 10,000 + 5,000 unfilled | Middle management, engineering managers | Year of Efficiency |
| February 2026 | ~7,500 | Recruiting, business ops, non-AI engineering | ~$1.5 billion H1 2026 charges |
| March 2026 | ~600 | Talent acquisition (recruiting) | Second 2026 wave |

*Source: Mark Zuckerberg public posts, Bloomberg, Reuters, Wall Street Journal, HRD Magazine*

### Why Does Meta Keep Cutting Recruiters?

The recruiter cuts reflect Meta’s view that growth will come from AI research hiring and that its broader talent acquisition function is oversized relative to remaining non-priority hiring needs. Specialist AI recruiting and executive search now absorb the workload that general-purpose recruiting once handled.

## The Year of Efficiency 2022-2023 Layoff Waves

- The November 2022 wave removed **11,000** employees and was Meta’s first mass layoff in company history. The November 2022 layoffs removed 11,000 employees; the cuts target middle management and product teams in the second mass layoff in five months.
- The March 2023 announcement framed the cuts as the Year of Efficiency, with plans to restructure teams and reduce the company by approximately **10,000** roles.
- The March 2023 wave, plus closed unfilled roles, totaled approximately **15,000** position reductions.
- Mark Zuckerberg framed the 2023 cuts as flattening the org structure and removing layers of middle management.
- The 2023 cuts targeted engineering managers and middle management whose work overlapped with other functions, per Bloomberg. The cuts target middle management and product teams whose work overlapped with other functions, with engineering managers a particular focus.
- FY2023 ended at **67,317** employees, the trough of the post-peak workforce. As of December 31, 2024.

| Wave | Date | Count | Function Focus |
|---|---|---|---|
| Wave 1 | November 9, 2022 | 11,000 | Recruiting, business partnerships, product |
| Wave 2 | March 14, 2023 | 10,000 + 5,000 unfilled | Mid-management, engineering managers |

*Source: Mark Zuckerberg Year of Efficiency announcement, Bloomberg, Meta annual report*

## The 10% Cut and AI Reallocation

- Meta announced on February 4, 2026, a workforce reduction of approximately **10%**, affecting roughly **7,500** roles. Meta Platforms will cut about 10% of its workforce; the cuts are expected to affect approximately 7,500 roles concentrated in recruiting, business operations, and engineering teams not directly building AI products.
- The cuts redirected compensation toward Meta Superintelligence Labs and AI infrastructure. Meta announced it will cut 10% of its workforce in 2026, redirecting saved compensation toward AI roles and Meta Superintelligence Labs hires.
- Cuts concentrated in recruiting, mid-management, and engineering functions, not directly building AI products. The reductions affect approximately 7,500 employees, with concentrations in recruiting, mid-management, and engineering functions not directly building AI products.
- AI research, AI infrastructure, and Reality Labs hiring continued through Q1 2026.
- Zuckerberg defended the cut on Threads as flattening management layers rather than contracting the company. We’re not getting smaller, Zuckerberg wrote, we’re getting flatter so that great individual contributors have more autonomy and AI tools amplify what each person can deliver.
- A second March 2026 wave added approximately **600** recruiter cuts. Meta is cutting roughly 600 recruiting roles in a second wave of reductions, deepening earlier February cuts that targeted about 7,500 roles.

| Category | Direction | Notes |
|---|---|---|
| Recruiting / Talent Acquisition | Down | Largest cut by share; March 2026 second wave |
| Mid-management | Down | Recurrence of 2023 Year of Efficiency pattern |
| Non-AI engineering | Down | Engineering teams not directly building AI |
| AI research | Up | MSL hiring continues |
| AI infrastructure | Up | Capex guide $115 to $135 billion FY2026 |
| Reality Labs | Up (hiring) | Continued investment despite $20.6 billion op-loss |

*Source: Reuters, Wall Street Journal, HRD Magazine, The Verge, Meta quarterly earnings release*

## Contractor and TVC Workforce Share

- Meta engages temporary employees, vendors, and contractors who supplement the full-time workforce on operational, content review, and infrastructure activities.
- The **76,834** fiscal-2025 figure represents full-time equivalents only and excludes TVC counts.
- TVC workforce is most heavily concentrated in content review, customer support, and infrastructure operations, per the 10-K language.
- The February 2026 reductions targeted full-time roles only; TVC pools were not specifically called out in the Reuters reporting on the cut. The cuts are expected to affect approximately 7,500 roles concentrated in recruiting, business operations, and engineering teams, not directly building AI products.
- Meta has not published a numeric TVC count in any FY2025 disclosure, leaving outside estimates to circulate without an official anchor. The 10-K language is unchanged from prior years on TVC disclosure; outside estimates from labor analysts and ex-vendor employees typically cite numbers in the 30,000-50,000 range for the global TVC pool, but those figures are not officially confirmed by Meta and should be treated as informed estimates rather than disclosed data.

| What 10-K Discloses | What’s Externally Estimated |
|---|---|
| 76,834 FTE at fiscal-2025 close | Numeric TVC count |
| “Temporary employees, vendors, and contractors” engaged | 30,000-50,000 TVC range (analyst estimates) |
| Operational, content review, infrastructure activities | Function-by-function TVC breakdown |

*Sources: SEC EDGAR (Meta 10-K Item 1)*

## Reality Labs Operating Loss vs Reality Labs Headcount

- RL operating loss was **$20.6 billion** in 2025, compared to **$17.7 billion** in 2024 and **$16.1 billion** in 2023. Reality Labs’ operating loss was $20.6 billion in 2025, compared to $17.7 billion in 2024 and $16.1 billion in 2023.
- RL employed approximately **12,000** people in 2025, per Reuters reporting. Meta’s Reality Labs division employs approximately 12,000 people across hardware engineering, software, and AI glasses operations.
- The implied operating loss per RL employee was approximately **$1.7 million** in 2025, per Reuters’ analysis cross-referenced with the FY2025 10-K. The implied operating loss per Reality Labs employee was approximately $1.7 million, the highest of any division at any major US tech company.
- RL revenue rose **48%** year over year to **$2.95 billion** in 2025 from **$2.0 billion** in 2024. Reality Labs revenue was $2.95 billion in 2025, an increase of 48% compared to $2.0 billion in 2024.
- Mark Zuckerberg told analysts in late January 2026 that 2026 Reality Labs losses will be similar to 2025 and will likely mark the peak. Mark Zuckerberg told analysts in late January 2026 that 2026 Reality Labs losses will be similar to 2025 and that this will likely mark the peak, after which losses will gradually decline.
- RL hires continued through Q1 2026 alongside the broader headcount reduction.

> **Key finding:** Per Reuters reporting cross-referenced with Meta’s FY2025 10-K Item 7 segment data, Reality Labs spent approximately **$1.7 million** in operating loss per RL employee in 2025, the highest such ratio at any major US tech company. Meta hires Reality Labs staff into a per-head burn higher than any FAANG peer pays in salary.

![Reality Labs Operating Loss Per Employee Trends](https://sqmagazine.co.uk/wp-content/uploads/2026/05/reality-labs-operating-loss-per-employee-trends.jpg "Reality Labs Operating Loss per Employee Trends")

## Menlo Park vs Global Office Footprint

- Meta operates in more than **40** cities across the Americas, Europe, the Middle East and Africa, and Asia Pacific. Meta has offices in more than 40 cities across the Americas, Europe, the Middle East and Africa, and Asia Pacific.
- Menlo Park is Meta’s global headquarters and largest engineering hub.
- Engineering and AI research are concentrated in Menlo Park, Seattle, Bellevue, New York, London, Tel Aviv, and Paris. Engineering and AI research are concentrated in Menlo Park, Seattle, Bellevue, New York, London, Tel Aviv, and Paris.
- Sales, marketing, and partnerships scale across regional hubs in London, Dublin, Singapore, and Tokyo. Sales, marketing, and partnerships scale across regional hubs in London, Dublin, Singapore, and Tokyo.
- Operations and customer support staff are concentrated in Austin and Dublin. Operations and customer support staff are concentrated in Austin and Dublin.
- Reality Labs engineering operates from Burlingame, Redmond, Sunnyvale, and Zurich. Reality Labs engineering operates from Burlingame, Redmond, Sunnyvale, and Zurich.

| Region | City | Functional Concentration |
|---|---|---|
| US West | Menlo Park, Seattle, Bellevue, Burlingame, Sunnyvale | HQ, engineering, AI, RL hardware |
| US East / Central | New York, Austin, Washington DC, Redmond | Engineering, ops, RL hardware, policy |
| EMEA | London, Dublin, Paris, Tel Aviv, Zurich | Engineering, AI, sales, RL |
| APAC | Singapore, Sydney, Tokyo | Sales, partnerships |

*Source: Meta Careers global office network*

## Revenue Per Employee and Productivity Metrics

- FY2025 revenue per employee was approximately **$2.62 million**, computed from **$200.97 billion** in revenue against **76,834** employees at fiscal-2025 close. Dividing FY2025 full-year revenue by the 10-K headcount yields the implied per-employee figure.
- FY2024 revenue per employee was approximately **$2.22 million**, computed from **$164.50 billion** against **74,067** employees. Dividing prior-year FY2024 revenue by the FY2024 10-K headcount yields the implied per-employee figure.
- Family of Apps generated **$193.7 billion** of the $200.97 billion total, concentrated in approximately 64,800 FoA staff. Family of Apps revenue was $193.7 billion.
- The RL revenue-per-employee figure was approximately **$246,000** in 2025, based on **$2.95 billion** in revenue across approximately **12,000** RL staff. Segment revenue of $2.95 billion divided by the 12,000 staff figure from Reuters yields the implied per-employee revenue figure.
- Meta’s overall revenue-per-employee figure sits at the upper end of FAANG productivity ratios. Comparable computations using public 10-K data put Apple’s revenue per employee in the $2.4 million range, Alphabet’s in the $1.9 million range, and Microsoft’s in the $1.0 million range. Meta’s $2.62 million is among the highest in big tech, with Nvidia being the standout outlier driven by its much smaller workforce.

| Year | Revenue | Headcount | Revenue per Employee (derived) |
|---|---|---|---|
| 2024 | $164.50 billion | 74,067 | ~$2.22 million |
| 2025 | $200.97 billion | 76,834 | ~$2.62 million |
| RL 2025 only | $2.95 billion | ~12,000 | ~$0.25 million |

*Sources: Meta Q4 earnings call, SEC EDGAR (Meta 10-K filings), Reuters*

## Common Questions

### Is Meta still hiring after the 2026 cuts?

Meta continued to hire in its highest priority areas during Q1 2026, including AI research, AI infrastructure, and Reality Labs, even as broader headcount fell. MSL hiring continues to outpace the rest of Meta even as the broader company sheds roles, per The Information. The pattern is reallocation, not contraction: roles in recruiting, mid-management, and non-AI engineering shrink while MSL and AI infrastructure expand.

### What is Meta’s revenue per employee compared to Google’s?

Meta’s full-year 2025 revenue was $200.97 billion, up 22% year over year from $164.50 billion in 2024. Dividing the full-year revenue by the 76,834 10-K headcount yields the implied per-employee productivity figure. Meta’s resulting **$2.62 million** per-employee figure compares against Google parent Alphabet at roughly $1.9 million on the same basis, putting Meta meaningfully ahead among FAANG peers on this productivity metric, though Nvidia leads big-tech overall on revenue-per-employee due to its smaller workforce.

## Conclusion

Meta’s workforce story reads as a reallocation, not a contraction. Meta ended 2025 at **76,834** employees per the 10-K disclosure. Headcount fell to approximately **74,700** in Q1 this year after the February reduction. The four layoff waves since November 2022 sum to approximately **29,100** roles trimmed, while net headcount sits roughly at the 2023 baseline. Meta Superintelligence Labs has grown to approximately **1,300** staff under Alexandr Wang’s first-year restructure. R&D consumed approximately **24%** of revenue at **$48.05 billion** in 2025, alongside a Reality Labs operating loss of **$20.6 billion**.

Meta’s FY2026 capital expenditure guide of **$115 billion to $135 billion** supports Meta Superintelligence Labs and core AI infrastructure. Tech-industry analysts and engineering-career planners watching the FAANG headcount picture should expect the layoff-to-AI-hire reallocation pattern to persist, with Reality Labs continuing to set the high-water mark for per-employee burn across approximately 12,000 staff and an approximately $1.7 million operating loss per RL employee.

Definition of AI Inference. Link to full glossary entry follows the description.**AI Inference**AI inference is the execution phase where a trained AI model applies what it learned to new, unseen data and produces an output such as a prediction.

[Read more](https://sqmagazine.co.uk/glossary/ai-inference/)