All companies
OpenAI

OpenAI

API-first
San Francisco, United StatesFounded 20157,000 employees$157B raised109 open roles

OpenAI is an artificial intelligence research and deployment company best known for ChatGPT and the GPT family of large language models. It offers consumer products (ChatGPT Free, Plus, Pro, Team, Enterprise), developer APIs, agentic coding tools (Codex), image generation, voice, and the Sora video model, as well as specialized models for scientific research.

AI mentor readresearched 1mo ago25 sources

OpenAI: $30B run-rate AI infrastructure giant with massive moat, but burning cash at scale and facing elite competition.

Strong contenderMedium conviction72/ 10072% confidence

A reasoned read from public sources. Each point links to its source.

Market & timing85
Product & moat72
Team62
Traction95
Competition45

The mentor's take

OpenAI is the most consequential AI company in the world right now, and the traction numbers are genuinely staggering — $25B ARR at 10x growth in two years is a once-in-a-generation revenue ramp. But the financial structure is deeply concerning: losses are scaling with revenue, the company is structurally dependent on infrastructure it doesn't own, and HSBC sees no path to profitability before 2030. The talent picture is mixed — Altman and Brockman are strong operators, but the departure of Sutskever and Karpathy (now at Anthropic) is a real signal about where the world's best AI researchers want to work. This is a company that could be the defining infrastructure of the AI era, or it could be a spectacular cash-burning cautionary tale if the model layer commoditizes before enterprise lock-in is achieved. The risk/reward is asymmetric but the downside is real.1234567

Market & timing

The LLM market is enormous and growing explosively. OpenAI alone targets a $700B+ TAM by 2030, and the broader AI infrastructure stack is absorbing over $160B in private capital across just three companies (OpenAI, Anthropic, xAI) in 18 months. ChatGPT has 900M+ weekly active users and 50M+ paying subscribers, with API volume above 15B tokens/min — numbers that suggest genuine mass-market penetration, not just hype. However, the market is structurally precarious: API costs dropped ~80% from 2025 to 2026, every major provider prices inference below cost to grab share, and HSBC estimates OpenAI won't be profitable by 2030 and needs another $207B in capital.86129

Product & moat

OpenAI has evolved well beyond a single model: it now ships ChatGPT (superapp with chat, coding, browsing), Atlas (AI-native browser), Codex (software engineering agent with 1.6M weekly users, tripling since start of 2026), the Agents SDK with a managed agent harness on AWS/Bedrock, and the Frontier enterprise platform. Greg Brockman explicitly states the company has 'line of sight to much better models this year' and believes text models will reach AGI. The product surface is expanding rapidly, but the company is simultaneously abandoning video generation and pivoting strategies — a sign of execution churn at the product layer.10311121314

Team

Sam Altman (CEO) is a seasoned operator with YC and startup pedigree, though he was ousted by the board in 2023 for lack of candor before being reinstated — a governance red flag that lingers. Greg Brockman (President) is deeply technical and actively shaping product strategy. Ilya Sutskever, who oversaw the research breakthroughs behind ChatGPT and o1, departed in mid-2024 to found Safe Superintelligence. Andrej Karpathy, a co-founder and deep learning luminary, left OpenAI in 2024 and joined Anthropic in 2026 — a notable talent drain to a direct competitor. The founding team's intellectual firepower has dispersed, though the company retains enormous institutional research depth.7151045

Traction

Traction is extraordinary by any measure. OpenAI hit $25B in annualized revenue by February 2026, up from $6B in 2024 and $2B in 2023 — roughly 10x in two years. Monthly revenue reached $2.5B in March 2026 (an 8x run-rate growth in 24 months). ChatGPT has 900M+ weekly active users, 50M+ paying subscribers, 9M+ paying business users, and 2.5B+ daily prompts. The company raised $122B in its latest round at an $852B valuation, with SoftBank ($30B), NVIDIA ($30B), Amazon ($50B), and Microsoft all participating. Codex weekly users tripled to 1.6M since January 2026.11617181419

Competition

Competition is fierce at every layer of the stack. At the model layer: Anthropic (Claude, $380B valuation, $19B ARR growing 1,167% YoY), Google/DeepMind (Gemini at 750M MAU), xAI ($230B valuation), Meta (open-source LLaMA), Mistral, and DeepSeek. At the platform/enterprise layer: Microsoft, AWS, Google Cloud, and Salesforce. At the product layer: Perplexity, GitHub Copilot, and others. Critically, Andrej Karpathy — one of the world's most respected AI researchers — joined Anthropic in 2026, signaling that talent competition is real. The open-source movement (Meta, Mistral, DeepSeek) structurally pressures API pricing and commoditizes model access.192021222365

The bull case

OpenAI has built the most powerful consumer distribution channel in AI history — 900M weekly users and 50M paying subscribers is a moat that is genuinely hard to replicate. The $122B raise at $852B valuation with strategic partners (NVIDIA, Amazon, SoftBank, Microsoft) creates a capital and compute advantage that compounds. Revenue is growing 10x in two years at a scale ($25B ARR) where most companies stall. The Agents SDK + Frontier enterprise platform + Codex + Atlas superapp strategy positions OpenAI to own the full AI workflow stack, not just the model layer. Enterprise penetration (9M+ paying business users, 75% of Fortune 500 via Microsoft integrations) creates durable switching costs.116171481213

The bear case

OpenAI lost $5B in 2024 on $3.7B in revenue and projects a $9B net loss in 2025 on $13B in sales — losses are scaling with revenue, not shrinking. HSBC estimates the company won't be profitable by 2030 and needs another $207B in capital beyond what it has raised. The company is structurally dependent on NVIDIA compute and Azure infrastructure it doesn't own, making it a 'tenant in a mansion built by its competitors.' Key founding researchers — Ilya Sutskever and Andrej Karpathy — have both departed, with Karpathy joining rival Anthropic in 2026. API prices dropped ~80% in a year, and open-source models from Meta, Mistral, and DeepSeek commoditize the model layer. The 2023 board crisis (Altman ousted for lack of candor) revealed governance fragility that hasn't been fully resolved.62345715

What would have to go right

OpenAI needs to convert its massive user base into durable, high-margin enterprise contracts before competitors commoditize the model layer further — the Frontier platform and Agents SDK are the bets here. The company must achieve compute self-sufficiency (via the $1.4T CapEx commitment referenced in analyst reports) to escape structural dependency on NVIDIA and Azure. Codex and the agent harness need to become the default developer infrastructure, not just a feature — the AWS/Bedrock partnership is a signal this is possible. Finally, the path to profitability requires that inference costs fall faster than revenue, which is plausible given the 80% API price drop trend but not guaranteed at the $9B+ annual loss run rate.391321412

Should you join?

If you're a senior big-tech engineer, OpenAI is one of the most impactful places you could work — the scale of deployment (900M weekly users, 15B+ tokens/min API volume) means your work has immediate, massive real-world effect. The mission is credible and the product surface is genuinely interesting. But be clear-eyed: this is not a typical equity-upside story. At an $852B valuation with $9B+ annual losses and no clear path to profitability before 2030, the equity math is hard. You're not getting in early — you're joining a late-stage company with public-company scale problems (org complexity, strategic churn, governance history) but without public-company liquidity. The comp will be strong, but the 10x equity outcome is unlikely at this valuation. Join if the mission and technical problems genuinely excite you; don't join primarily for financial upside.1162171045

Comp
Cash comp will be strong (big-tech competitive or above). RSUs at $852B valuation have limited upside unless you believe in a multi-trillion outcome. Pre-IPO liquidity is uncertain despite IPO talk.
Stage vs equity
At $852B valuation with ongoing losses, equity upside is severely compressed. A 3-5x outcome (to $2.5-4T market cap) is possible but requires OpenAI to become one of the most valuable companies in history AND achieve profitability — a high bar. Comp will be competitive but don't model life-changing equity.
Who you'd work with
You'd work alongside a large, institutionally deep research and engineering org. The founding luminaries (Sutskever, Karpathy) have left, but the company retains significant talent density. Greg Brockman is actively technical and product-focused.

To watch

  • 01Monthly revenue trajectory: does the $2.5B/month run rate continue to grow, or does it plateau as API prices collapse and competition intensifies?
  • 02Path to profitability: does the $9B projected 2025 net loss shrink in 2026, or do losses continue to scale with revenue?
  • 03Enterprise lock-in: does the Frontier platform and Agents SDK achieve genuine switching costs, or do enterprises multi-vendor across OpenAI, Anthropic, and Google?
  • 04Talent retention: do more senior researchers follow Sutskever and Karpathy out the door, particularly to Anthropic or SSI?
  • 05Compute independence: does OpenAI make meaningful progress on its $1.4T CapEx commitment to own its own infrastructure, reducing NVIDIA/Azure dependency?

Key risks

  • 01Structural unprofitability: $9B+ projected net loss in 2025 with HSBC estimating no profitability before 2030 and $207B more capital needed — the cash burn is existential if revenue growth slows.
  • 02Model layer commoditization: API prices dropped ~80% in one year; open-source models (Meta LLaMA, Mistral, DeepSeek) threaten to make frontier model access a near-zero-cost commodity.
  • 03Infrastructure dependency: OpenAI is structurally dependent on NVIDIA compute and Azure cloud it doesn't own, giving its direct competitors (Microsoft, Google) leverage over its cost structure.
  • 04Elite talent atrophy: Ilya Sutskever (chief scientist, led ChatGPT and o1 research) and Andrej Karpathy (co-founder, deep learning luminary) have both departed — Karpathy to Anthropic — raising questions about sustained research leadership.
  • 05Governance fragility: the 2023 board crisis (Altman ousted for 'not being consistently candid') and the complex PBC/nonprofit structure create ongoing governance risk, especially pre-IPO.

Sources

  1. 1OpenAI revenue, valuation & funding | Sacra·sacra.com
  2. 2OpenAI won’t make money by 2030 and still needs to come up with another $207 billion to power its growth plans, HSBC estimates | Fortune·fortune.com
  3. 3Open AI's Strategy - A detailed teardown·apoorvamittal.substack.com
  4. 4Ilya Sutskever - Wikipedia·en.wikipedia.org
  5. 5Andrej Karpathy·en.wikipedia.org
  6. 6Large Language Model Market 2026: Revenue & Race·analysis-atlas.com
  7. 7Sam Altman·en.wikipedia.org
  8. 8OpenAI's $20B ARR: Mapping the Scalability Flywheel to $700B TAM·ainvest.com
  9. 9Deep|OpenAI: How Is OpenAI Eating the Enterprise SaaS TAM?·fundaai.substack.com
  10. 10OpenAI President Greg Brockman: Doubling Down on Text Models, The Superapp Plan, Codex’s Potential·bigtechnology.com
  11. 11How we built OWL, the new architecture behind our ChatGPT-based browser, Atlas | OpenAI·openai.com
  12. 12Introducing OpenAI Frontier | OpenAI·openai.com
  13. 13The Runtime Is the Moat: OpenAI’s New Agent Harness, on AWS·pumpingcode.substack.com
  14. 14Scaling AI for everyone - OpenAI·openai.com
  15. 15OpenAI - Wikipedia·en.wikipedia.org
  16. 16OpenAI's $30 Billion Revenue Run Rate Makes the IPO Thesis Undeniable - Networkcraft·networkcraft.net
  17. 17OpenAI raises $122 billion to accelerate the next phase of AI·openai.com
  18. 18OpenAI secures an extra $10 billion in record funding round, CFO Friar says·cnbc.com
  19. 19Top OpenAI Alternatives, Competitors·cbinsights.com
  20. 20OpenAI Competitors: Who’s Really Competing and How - Hot Shot·hotshot.co
  21. 21OpenAI Competitors: Complete List & Market Landscape·distillintelligence.com
  22. 2214 OpenAI Competitors and Alternatives As Of 2026 - RankRed·rankred.com
  23. 23Top OpenAI Competitors & Alternatives 2026 | Gartner Peer Insights - Generative AI Knowledge Management Apps/General Productivity·gartner.com

About

OpenAI is an artificial intelligence research and deployment company best known for ChatGPT and the GPT family of large language models. It offers consumer products (ChatGPT Free, Plus, Pro, Team, Enterprise), developer APIs, agentic coding tools (Codex), image generation, voice, and the Sora video model, as well as specialized models for scientific research.

OpenAI was founded in December 2015 as a non-profit AI research lab by Sam Altman, Elon Musk, Greg Brockman, Ilya Sutskever, Wojciech Zaremba, John Schulman and others, with the mission of ensuring artificial general intelligence benefits all of humanity. It restructured to a capped-profit model in 2019 to attract the capital required to train frontier models.

Recently

8 updates

Founders

SA

Sam Altman

Co-Founder & CEO

Co-founder and CEO of OpenAI; former president of Y Combinator (2014-2019).

GB

Greg Brockman

Co-Founder & President

Co-founder and President of OpenAI; previously CTO at Stripe.

IS

Ilya Sutskever

Co-Founder (former Chief Scientist)

Co-founder and former Chief Scientist of OpenAI; world expert in deep learning. Departed in 2024 to start Safe Superintelligence Inc.

EM

Elon Musk

Co-Founder (departed 2018)

Co-founder and early funder of OpenAI; departed the board in 2018.

WZ

Wojciech Zaremba

Co-Founder

Co-founder of OpenAI, leads the Codex / coding research team.

JS

John Schulman

Co-Founder (departed 2024)

Co-founder of OpenAI and leading reinforcement-learning researcher; departed in 2024 for Anthropic.

Funding

$157B raised total

H1B visa sponsorship

Source: USCIS

Petitioner on record

OPENAI OPCO LLC · SAN FRANCISCO, CA

Approvals
76
Success rate
98.7%
New hires
9
Denials
1