01
Consumer entry point
Free and paid ChatGPT establish global habit; advertising begins to monetize free traffic.
OPENAIINTELLIGENCE STACK · MODELS, PRODUCTS & DISTRIBUTION
OpenAI · The Intelligence Layer
Model capability must become durable use and unit economics through products, developer interfaces and distribution.
INTELLIGENCE STACK
Turn capability into products and distribution
Models, user products and platform interfaces each carry a different commercialization task.
- Stage 1
Frontier Models
Reasoning, multimodality and the capability frontier.
- Stage 2
ChatGPT & Codex
Entry points for conversation, knowledge work and software agents.
- Stage 3
API & Agents
Responses API, tools, evaluations and developer distribution.
All companies
Private companyDeep dive completeData as of 2026-08-03
OpenAI
OpenAI Group PBC
Distribution and revenue growth are proven. Price and capital efficiency are not.
900M weekly users and a $24B revenue run rate prove global distribution, while the $852B valuation demands growth, margin and compute efficiency at once.
Next mandatory review2026-09-01
$852B
Post-money valuation
Company disclosed, 2026-03-31
$24B
Annualized revenue run rate
Annualized from $2B monthly revenue
35.5×
Valuation / run rate
Calculated from disclosed figures
>900M
ChatGPT weekly users
More than 50M subscribers
01 · Economic engine
From model capability to workflow.
02
Enterprise and API
Enterprise exceeds 40% of revenue, but retention, channel take and segment margin remain undisclosed.
03
Codex workflow
Agents enter repositories, permissions and approvals, determining whether distribution becomes a high-switching-cost platform.
02 · Valuation threshold
What the private price demands.
Threshold analysis, not a target priceAnchorRequired economicsFrom current run rate
10× revenue$85.2B revenueRequired growth88% two-year CAGR
8× revenue$106.5B revenueRequired growth111% two-year CAGR
6× revenue$142B revenueRequired growth143% two-year CAGR
30× free cash flow$28.4B free cash flowRequired growthNot disclosed
03 · The argument
Strongest bull case
ChatGPT has rare global consumer distribution alongside enterprise, developer, coding-agent and advertising monetization. It could become an operating-system-level AI platform.
Strongest bear case
Model prices keep falling, Microsoft revenue sharing runs through 2030, and cloud and chip partners may capture much of the profit while compute commitments dwarf current revenue.
Current ruling
The bull case wins on distribution and growth; the bear case wins on price and capital efficiency. Before the public S-1, $852B is not a verifiable intrinsic value.
04 · Falsification
What would change the judgment.
- 01The S-1 reveals a large gap between monthly run-rate and GAAP revenue, or reliance on one-offs and gross presentation.
- 02Compute contracts contain rigid minimums while utilization, pricing or user growth slows.
- 03Codex paid expansion falls, enterprises adopt model routing systematically, and ChatGPT retention weakens.