--- title: "How Many People Work at OpenAI 2026: Headcount, Growth & Hiring Plan" date: 2025-08-18 author: "Sofia Ramirez" featured_image: "https://sqmagazine.co.uk/wp-content/uploads/2025/08/how-many-people-work-at-openai.jpg" categories: - name: "Artificial Intelligence" url: "/artificial-intelligence.md" tags: - name: "Statistics" url: "/tag/statistics.md" --- # How Many People Work at OpenAI 2026: Headcount, Growth & Hiring Plan OpenAI employs somewhere between 4,500 and 8,000 people, with the headline figure depending on which methodology you trust. Revelio Labs workforce intelligence put OpenAI Foundation at 7,850 employees in December 2025, while the Financial Times reported in March 2026 that OpenAI’s headcount sat near 4,500 with a plan to reach 8,000 by year-end. The gap is methodology, not contradiction, and the data inside it tells the real story of how fast [OpenAI statistics](https://sqmagazine.co.uk/openai-statistics/) has scaled. The triangulation matters because most workforce pages pick one and run it as definitive. ## Key Takeaways - OpenAI Foundation reported **7,850** employees as of December 2025, a **54.9%** year-over-year growth rate from 4,467 at end-2024. - The Financial Times reported in March 2026 that OpenAI planned to grow from about **4,500** to **8,000** by the end of 2026, a near-doubling. - Engineering is OpenAI’s largest function at about **56%** of total headcount, with research at roughly **25%**. - Epoch AI’s May 8, 2026 analysis put OpenAI’s revenue per employee at an estimated **$5.5 million**, behind Anthropic’s estimated **$9 million** but well above Nvidia’s **$5.1 million**. - Median [software engineer total compensation](https://sqmagazine.co.uk/software-engineer-layoff-statistics/) at OpenAI was **$555,000**, ranging from **$249,000** at L2 to **$1.28 million** at L6, per Levels.fyi data. - Sam Altman told staff on **January 27, 2026,** that OpenAI planned to “dramatically slow down” hiring, 8 weeks before the FT report on the 8,000 target. - After the 2024 departure wave, only **3 of 11** of OpenAI’s original co-founders remain active at the company. ## Editor’s Choice: How Many People Work at OpenAI? - **7,850 employees** is the December 2025 Revelio Labs headline figure for OpenAI Foundation’s global workforce. - **4,500 employees** is the March 2026 Financial Times figure for OpenAI’s direct payroll, citing two unnamed sources with direct knowledge. - **8,000 employees** is the target headcount OpenAI was reported to be pursuing by the end of 2026. - An estimated **$5.5 million** per employee in revenue is the Epoch AI May 2026 productivity figure for OpenAI. - **$1.5 million** was the average stock-based compensation per OpenAI worker in 2025. - **1,480 engineers** is the implied engineering headcount at the 4,500-base, **56%** mix. - **26%** Foundation / **27%** Microsoft / **47%** employees and investors is the post-restructuring ownership split since October 28, 2025. ## Recent Developments - **March 21, 2026**: The Financial Times reported OpenAI planning to grow from **4,500 to 8,000** employees by the end of 2026, with new hiring concentrated in product development, engineering, research, and sales. - **January 27, 2026**: Sam Altman told a livestreamed town hall that OpenAI planned to “dramatically slow down how quickly we grow because we think we’ll be able to do so much more with fewer people.” - **Q1 2026**: OpenAI’s Annual Recurring Revenue reached **$12.7 billion**, a **150%** increase over Q1 2025, supporting the case for aggressive hiring even as Altman flagged restraint. - **May 8, 2026**: Epoch AI published a revenue-per-employee comparison putting OpenAI at an estimated $5.5 million and Anthropic at an estimated **$9 million**, ahead of every public tech company on the Forbes Global 2000. - **October 28, 2025**: OpenAI restructured into a public benefit corporation, with the OpenAI Group PBC owned **26%** by the OpenAI Foundation, **27%** by Microsoft, and **47%** by employees and investors. ## How Many Employees Does OpenAI Have OpenAI’s 2026 headcount sits between 4,500 and 7,850 depending on the data source and the measurement window. Revelio Labs, which standardizes LinkedIn-public employment records, reported **7,850 employees** at OpenAI Foundation in its December 2025 snapshot, a **54.9%** year-over-year increase from 4,467 at end-2024. The Financial Times reported in March 2026 that OpenAI’s payroll was nearer **4,500** with a target of **8,000** by year-end, citing two unnamed sources with direct knowledge. The methodology delta, LinkedIn-public scrape versus direct-payroll reporting, accounts for most of the gap. SourceReported headcountAs ofMethodRevelio Labs7,850December 2025LinkedIn-public employment recordsFinancial Times (via CNBC)4,500March 2026Two unnamed sources with direct knowledgeWikipedia infobox (corporate disclosure)4,5002026Public filing summaryOpenAI target plan (FT-reported)8,000End of 2026Stated hiring target*Source: Revelio Labs, Financial Times via CNBC, Wikipedia, OpenAI hiring plan* OpenAI Foundation employees (December 2025) VALUE · Source: Source: Revelio Labs December 2025 snapshot VALUE · SQ MAGAZINE SNAPSHOT OpenAI Foundation employees (December 2025) Revelio Labs · 2025 7,850 SOURCE Source: Revelio Labs December 2025 snapshot ## OpenAI Headcount Growth OpenAI’s headcount expanded roughly 1,300x between its 2015 founding and end of 2025. OpenAI was founded on December 8, 2015, as a nonprofit organization. Headcount grew from 6 in 2015 to 335 in 2022, then more than doubled to 770 in 2023 as [ChatGPT](https://sqmagazine.co.uk/chatgpt-statistics/) shipped and demand exploded. Revelio Labs put end-2023 headcount at 2,932, end-2024 at 4,467, and end-2025 at 7,849, with a 31.6% jump in 2023, a 41.5% jump in 2024, and a 54.9% jump in 2025. - 2015: **6** employees at founding - 2022: **335** employees - 2023: **770** employees (more than double 2022) - End-2023: **2,932** (+31.6% on the Revelio Labs LinkedIn-public series) - End-2024: **4,467** (+41.5%) - End-2025: **7,849** (+54.9% year-over-year) - March 2026: **4,500** per the Financial Times direct-source reporting - End-2026 target: **8,000** per the Financial Times reporting Year by Headcount (Revelio Labs series) HEADCOUNT (REVELIO LABS SERIES) · Headcount (Revelio Labs series) · Source: Source: Revelio Labs December 2025 snapshot; founding and early figures from MakerStations aggregation of OpenAI corporate disclosures HEADCOUNT (REVELIO LABS SERIES) · SQ MAGAZINE ANALYSIS Year by Headcount (Revelio Labs series) Headcount (Revelio Labs series) Revelio Labs · 2025 8000 6400 4800 3200 1600 0 2015: 2015 2022: 2022 2023: 2023 2024: 2024 2025: 2025 2015 2022 2023 2024 2025 SOURCE Source: Revelio Labs December 2025 snapshot; founding and early figures from MakerStations aggregation of OpenAI corporate disclosures > **By the numbers:** Revelio Labs recorded a **54.9%** year-over-year increase in 2025, on top of **41.5%** in 2024 and **31.6%** in 2023. The compound effect across the three years is large enough that any quarterly snapshot lands stale within months. ## OpenAI Employees by Department and Function Engineering dominates the OpenAI workforce, and research follows well behind it. Engineering is OpenAI’s largest department at roughly **56%** of total headcount, with research at about **25%**. The remaining **19%** spreads across commercial, support, and administrative functions. At the 4,500-employee FT base, that implies about 1,480 engineers. - **Engineering:** **~56%** of total headcount - **Research:** **~25%** of total headcount - Commercial, sales, support, and administrative roles fill the remainder, with the rest split across those functions. - **2026 hiring focus per the Financial Times:** product development, engineering, research, and sales, plus “specialists for technical ambassadorship” who help businesses use OpenAI’s tools FunctionShare of headcountEngineering56%Research25%Commercial, support, adminRemainder*Source: Revelio Labs role breakdown via MakerStations; Financial Times via CNBC for the 2026 hiring focus* The engineering share is high for this revenue scale. The shape reflects OpenAI’s research-product fusion. ## OpenAI Office Locations and Geographic Workforce Footprint OpenAI runs offices across San Francisco, NYC, Seattle, London, Dublin, Paris, Brussels, Singapore, and Tokyo. Headquarters is located at 1455 3rd Street, San Francisco, California. We’re opening offices in NYC, Seattle, Paris, Brussels, and Singapore, alongside our growing San Francisco, London, Dublin, and Tokyo offices. The San Francisco footprint has expanded sharply. OpenAI secured additional San Francisco office space as of March 2026, expanding its footprint to more than 1 million square feet, per the Financial Times. Revelio Labs’ end-2024 OpenAI Foundation count was roughly **4,467** employees across all locations. - San Francisco (HQ), London, Dublin, and Tokyo were the four established offices before the October 2024 expansion - NYC, Seattle, Paris, Brussels, and Singapore were added as new offices in the October 2024 announcement - More than **1 million square feet** of San Francisco real estate as of early 2026 ## OpenAI’s Hiring Plan and the Altman Tension OpenAI’s stated hiring trajectory carries an unusual internal conflict. The Financial Times reported on March 21, 2026, that OpenAI was looking to expand its workforce to **8,000** employees by the end of 2026, nearly doubling staff from its current headcount of **4,500**. New positions were said to span product development, engineering, research, and sales, with a new “technical ambassadorship” specialty for client-facing roles. But the same Sam Altman who signed off on that growth plan had told staff We are planning to dramatically slow down how quickly we grow because we think we’ll be able to do so much more with fewer people. That town hall ran on January 27, 2026, about 8 weeks before the FT plan surfaced. Altman framed it as restraint, not a freeze: So I think the right approach for us will be to hire more slowly but keep hiring. - Stated end-2026 target: **8,000 employees** - March 2026 base: **4,500 employees** - Altman’s January 27, 2026 framing: “dramatically slow down” - Stated hiring focus: product development, engineering, research, sales, technical ambassadorship - Competitive pressure: AI Index data cited by Engadget showed businesses were 70% more likely to go with Anthropic when buying AI services for the first time > **Why it matters:** The 8-week gap between Altman’s slowdown framing and the FT plan is the tension worth tracking. The stated target requires hiring roughly **3,500** people in 9 months. **The likely resolution:** The FT figure capped a planning range while Altman’s slowdown signaled internal calibration. Both can be true; only one lands at year-end. ## OpenAI Revenue Per Employee Compared to Big Tech OpenAI’s revenue per employee places it in a different league from any public tech peer. Epoch AI’s May 8, 2026 analysis put OpenAI at an estimated **$5.5 million** in revenue per employee, with Anthropic at **$9 million**. By comparison, Nvidia sits at **$5.1 million** per 10-K filings, the highest of any large-cap public tech company. Company by Revenue per employee REVENUE PER EMPLOYEE · Revenue per employee (USD) · Source: Source: Epoch AI Data Insights (May 8, 2026); mlq.ai derivation from public ARR statements REVENUE PER EMPLOYEE · SQ MAGAZINE ANALYSIS Company by Revenue per employee Revenue per employee (USD) Epoch AI · 2026 10M 7.5M 5M 2.5M 0 ~$9.0M Anthropic ~$5.5M OpenAI (Epoch AI) ~$5.1M Nvidia ~$3.3M OpenAI (mlq.ai Q1 2026) ~$1.8M Google ~$1.6M Meta SOURCE Source: Epoch AI Data Insights (May 8, 2026); mlq.ai derivation from public ARR statements A different cut tells the same story. OpenAI revenue per employee at Q1 2026 was approximately **$3.3 million**, compared to Google’s **$1.8 million** and Meta’s **$1.6 million**. Company by Revenue per employee ($M) REVENUE PER EMPLOYEE ($M) · Revenue per employee ($M) · Source: Epoch AI Data Insights, May 2026 REVENUE PER EMPLOYEE ($M) · SQ MAGAZINE ANALYSIS Company by Revenue per employee ($M) Revenue per employee ($M) Epoch AI · 2026 10 7.5 5 2.5 0 9.0 Anthropic 5.5 OpenAI (Epoch) 5.1 Nvidia 1.8 Google 1.6 Meta SOURCE Source: Epoch AI Data Insights, May 2026 **Two figures, one story:** Epoch uses run-rate revenue and a smaller headcount; mlq.ai uses ARR over Revelio’s count. Both show the productivity gap that funds engineer comp. ## OpenAI Compensation and Engineer Salaries OpenAI’s pay scales are at the top of the tech industry. Median software engineer total compensation is roughly $555,000 per year, with the L2 to L6 range running from $249,000 to $1.28 million. Research scientists earn more, with L4 to L5 total compensation between $771,000 and $1.47 million-plus. - Software Engineer (Member of Technical Staff): **$170,000 to up to $490,000** - Hardware Engineer: up to **$555,000** - Data Science: up to **$490,000** - Research Scientist L4 to L5: **$771,000 to $1.47 million** - Average stock-based compensation in 2025: **$1.5 million** per worker - Sam Altman base salary in 2024 per federal filing: **$76,001** - Meta has reportedly offered OpenAI talent packages worth upwards of **$100 million** Role band by Total compensation range TOTAL COMPENSATION RANGE · Total compensation range (USD) · Source: Source: Levels.fyi via Pin and JobsByCulture (March 2026) TOTAL COMPENSATION RANGE · SQ MAGAZINE ANALYSIS Role band by Total compensation range Total compensation range (USD) Levels.fyi · 2026 Software Engineer L2 ~$249,000 Software Engineer median ~$555,000 Software Engineer L6 ~$1.28M Research Scientist L4 ~$771,000 0 250K 500K 750K 1M 1.2M 1.5M SOURCE Source: Levels.fyi via Pin and JobsByCulture (March 2026) > **Worth noting:** Altman’s **$76,001** disclosed 2024 base salary stands out against engineer medians of **$555,000** and research scientist top bands above **$1.47 million**. The CEO’s pay is structured around equity rather than cash, which keeps Big Tech CEO-to-worker ratios off the proxy filings. ## OpenAI Co-Founder and Executive Departures The 2024 departure wave saw two co-founders and senior research leadership leave OpenAI. DateDeparting executiveRoleDestinationMay 2024Ilya SutskeverCo-founder, Chief ScientistSafe Superintelligence Inc.2024John SchulmanCo-founderAnthropicSeptember 2024Mira MuratiCTOIndependent ventureSeptember 2024Barret ZophVP of ResearchIndependentSeptember 2024Bob McGrewChief Research OfficerIndependent*Source: Express Tribune aggregation of public departure announcements via X* Co-founder Ilya Sutskever announced his departure in **May 2024** and went on to found Safe Superintelligence Inc. Co-founder John Schulman said he was leaving in **2024** to join rival Anthropic. CTO Mira Murati announced her departure in late September 2024, and within hours, VP of Research Barret Zoph and Chief Research Officer Bob McGrew also announced they were leaving. **The net effect:** **only 3** of OpenAI’s original 11 co-founders remain active at the company. The remaining active co-founders include Sam Altman (CEO) and Greg Brockman (President), both still listed on the leadership team after the October 28, 2025 restructuring. See our peer headcount coverage at [Google headcount](https://sqmagazine.co.uk/how-many-people-work-at-google/). The departure pattern points at a structural strain that the 2024 window concentrated rather than caused, with the same-day senior-research exits reading as coordinated calibration. ## OpenAI vs Anthropic Workforce Comparison Anthropic is the most useful comparator because it’s the only AI lab generating revenue at scale with a similar product mix. Anthropic had between **3,000 and 5,000 employees** as of March 2026, with Tracxn estimating about 5,000. Anthropic’s annualized run rate climbed from **$1 billion** at end-2024 to more than **$30 billion** by end of March 2026, implying revenue per employee in the $6 million to $10 million band. Lab2026 headcountAnnualized revenueImplied RPEOpenAI (FT)~4,500~$24 billion~$5.3 millionOpenAI (Revelio)~7,850~$25 billion~$3.2 millionAnthropic (Tracxn)~5,000~$30 billion~$6.0 million*Source: Revelio Labs, Tracxn via SaaStr, Financial Times, OpenAI ARR statements* Lab by Headcount HEADCOUNT · Headcount · Source: Revelio Labs, Financial Times, Tracxn via SaaStr HEADCOUNT · SQ MAGAZINE ANALYSIS Lab by Headcount Headcount Revelio Labs OpenAI (Revelio Labs, Dec 2025) 7,850 Anthropic (Tracxn, March 2026) 5,000 OpenAI (FT direct, March 2026) 4,500 0 1333 2666 4000 5333 6666 8000 SOURCE Source: Revelio Labs, Financial Times, Tracxn via SaaStr At a more than **$30 billion** revenue threshold, historical comparisons show Google had 32,000 employees, and Salesforce had 79,000 employees, while Anthropic reaches that revenue line with about 5,000. > **Key finding:** Epoch AI’s May 2026 analysis, both Anthropic at an estimated **$9 million** and OpenAI at an estimated **$5.5 million** per employee would outperform every public tech company on Forbes’ Global 2000 list if either company went public. Nvidia at $5.1 million is the public ceiling, and Anthropic clears it by a wide margin. ## OpenAI’s Ownership Structure and Headcount Implications OpenAI restructured into a public benefit corporation on October 28, 2025, with the for-profit branch renamed OpenAI Group PBC and the nonprofit renamed OpenAI Foundation. The post-conversion ownership split: OpenAI Foundation holds **26%**, [Microsoft](https://sqmagazine.co.uk/microsoft-statistics/) holds **27%**, and employees plus investors hold **47%**. All but one of the board members of the OpenAI Foundation were also board members of OpenAI Group PBC as of April 2026, with Bret Taylor chairing the Foundation board and Sarah Friar serving as CFO. - OpenAI Foundation: **26%** (nonprofit, parent) - Microsoft: **27%** (major external holder) - Employees plus investors: **47%** (aggregate bloc) - Headquarters: 1455 3rd Street, San Francisco - Equity vesting and stock-based compensation of **$1.5 million** per worker average (2025) places employees on the receiving end of the largest bloc by share ## Is OpenAI bigger than Google? OpenAI is far smaller than Google by headcount but far larger on a revenue-per-employee basis. At Q1 2026, OpenAI’s revenue per employee was about **$3.3 million** versus Google’s **$1.8 million**, so OpenAI generates roughly 1.8x the revenue per headcount. By total headcount, Google had around **32,000 employees** at the more than **$30 billion** revenue mark for historical comparison, while OpenAI sits between 4,500 and 7,850. Different definitions of “bigger” give different answers: productivity favors OpenAI; workforce scale favors Google. ## Who owns 51% of OpenAI? No single party owns 51% of OpenAI under the current structure. Since restructuring on October 28, 2025, OpenAI Foundation holds **26%**, Microsoft holds **27%**, and employees plus investors hold **47%** of OpenAI Group PBC, with the Foundation board appointing and able to remove PBC board members. The largest single bloc is employees plus investors, then Microsoft, then the Foundation. The single-majority framing predates the PBC conversion. ## Is OpenAI losing $14 billion? Public reporting suggests OpenAI is operating at a substantial loss; no single confirmed “$14 billion” loss figure appears in primary disclosures. Altman’s January 27, 2026 town hall coverage said OpenAI is losing billions of dollars each quarter, with a cash burn rate described as “immense” and plans to spend over **$1 trillion** on data center infrastructure over time. The financial pressure context informs the “code red” memo Altman issued to staff the previous month and his subsequent January 27, 2026 framing on hiring restraint. Specific loss figures vary across reports; the qualitative picture is multi-billion-dollar quarterly losses against the ARR captured above. ## Conclusion OpenAI employs between 4,500 and 7,850 people, with engineering at roughly 56% of staff and revenue per employee in the approximately **$3.3 million** to **$5.5 million** range. The range bookends a methodology problem, and the FT’s year-end target sits in structural tension with Altman’s January framing on hiring restraint. Whatever year-end headcount lands at, the productivity story is the one to watch: our [Claude vs ChatGPT statistics](https://sqmagazine.co.uk/claude-vs-chatgpt-statistics/) point at output-per-person as the metric reshaping every tech company’s hiring calculus.