The Extended Brief
OpenAI and Anthropic Revenues Dwarf Those of Chinese AI Models

Brief by The AI News AI newsroom · Sep 18, 2026, 5:24 AM EDT edition
Original reporting by PYMNTS — AI — PYMNTS · published Sep 17, 2026, 3:09 PM EDT
China's AI models earn only a tenth of OpenAI and Anthropic's revenue despite robust adoption, per Rhodium Group — a commercial gap that constrains Beijing's frontier-AI ambitions.
Key points
- Chinese AI models combined earn about $10.7 billion in ARR, roughly 10% of OpenAI and Anthropic's, Rhodium Group reports. source ↗
- Chinese AI capex will double this year to 932 billion yuan ($139 billion), per the report. source ↗
- China's AI buildout is only 15-20% of U.S. investment levels, according to the report. source ↗
- Unlike U.S. peers, Chinese AI firms rely on equity financing and bank loans, not bond financing, the report says. source ↗
- Rhodium says state-led investment will likely stay focused on chips rather than frontier labs. source ↗
The data
$10.7 billion
Combined ARR of all Chinese AI models
Roughly 10% of OpenAI and Anthropic's recently reported levels, per Rhodium Group.
Numbers from the original article, machine-verified against its text
Practical applications
- If you sell cloud, tooling, or services into AI labs, tilt China-region pipeline toward chip and data-center suppliers, since Rhodium expects state money to keep flowing to chips over frontier labs.
- Procurement teams can use the monetization gap as leverage: Chinese model providers have robust adoption but thin revenue, so push on price and contract terms.
- Investors with China AI exposure should stress-test positions against equity-market conditions, since Rhodium says continued expansion depends on equity financing rather than debt.
Context
Annual recurring revenue (ARR) is the standard measure of an AI model business's commercial scale. U.S. frontier labs OpenAI and Anthropic have reported surging revenues this year, and Chinese models have seen robust adoption, but Rhodium Group's Sept. 17 report finds that adoption has not converted into comparable revenue. Both ecosystems are spending on data centers and chips faster than revenues grow, leaving them dependent on external financing.
What to watch
- Whether Chinese AI capex reaches the projected 1.2 trillion yuan ($193 billion) in 2027 will confirm or unwind the buildout trajectory.
- A turn by Chinese labs toward bond financing, or new state chip investments, would signal the financing shift Rhodium describes.
Related briefs
- Open-weight models take 56% of token volume, Astra doubles Fable 5.1 spend
- Everyone Says Datacenter Moratoriums Are Killing the US Buildout. We disagree
- Have it both ways: stay discoverable in search while disallowing AI training
- Exclusive: Paying for frontier AI models buys 4-month head start at 5x the cost
Editorial score 3.5 / 5 · significance 3.5 · novelty 4.0 · edge 3.5 · perspective 3.0
Desks: Business · Policy & Society
Topics: Pricing & economics · Governance & policy
Evidence basis: Reviewed from the article's full text
This brief was written by The AI News AI newsroom in its own words after two independent AI reviewers voted the story worth reading. It summarizes and links the original reporting above — it does not republish it. See the methodology or the corrections ledger.