I don’t understand the comments about this business failing. Objectively, this has the fundamentals to be an incredible business.
- Burn $8.5B to add $6B in top line.
- Rule of 40 is right around 40. With a product that requires huge capex and has incredible gross margins long term.
- 1B weekly active users gives them n of 1 distribution to expand ARPU through additional products and services.
Is $500B crazy expensive? Yes. Will this company be worth trillions one day? More likely yes than no based off this update.
OpenAI's financial situation (via The Information)
•Revenue: $4.3B in H1 2025 (+16% vs all of 2024).
•Losses: Operating loss $7.8B, net loss $13.5B (over half from remeasurement of convertible equity).
•Cash burn: $2.5B in H1 2025; projected $8.5B for full year.
•R&D: Biggest expense, $6.7B in H1 2025 (on track to double vs 2024).
•Stock compensation: ~$2.5B in H1 (nearly doubled YoY); targeting ~$6B for full year.
•Sales & marketing: $2B in H1 (almost double all of 2024; includes a Super Bowl ad).
•Microsoft: 20% revenue share; projected to save $50B by 2030 as this percentage declines.
•Server costs: $2.5B cost of revenue in H1, mainly renting Nvidia-powered servers via Microsoft.
•Cash reserves: $17.5B in cash + securities at end of Q2; +$10B new funding in June; seeking +$30B in July.
•Valuation: Tender offer values OpenAI at ~$500B (vs $260B at start of 2025).
•Long-term projections: $200B revenue in 2030, $115B cash burn until then, $450B server costs (mostly rented).
•Nvidia deal: LOI for $10B per 1 GW data center capacity (up to $100B).
•User targets: 1B weekly ChatGPT users by end of 2025 (vs 500M in March).
In short: explosive revenue growth, but even higher losses and cash burn, driven by R&D, server costs, and stock comp. Valuation is soaring, while the roadmap assumes massive investment and scale.