40 curated ecosystem entities

China's Frontier AI Ecosystem

China's AI ecosystem mapped across capital, compute and data infrastructure, institutional linkages, and commercialization. The analytical lens draws on developmental orchestration and is anchored in Beijing's AI Pilot Zone, with comparative ecosystem references across China.

  • 40Ecosystem entities
  • 5City clusters
  • 6Sectors
  • 3Orchestration domains

China AI Ecosystem Table

Evidence protocol. Core table fields are manually verified against dated public sources and frozen as of 17 September 2026. The optional news-signal feed is refreshed separately for context and does not independently verify financial, ownership, policy-linkage, or relationship claims.

Entity ↕ Entity type ↕ City ↕ Case role ↕ Segment ↕ Founded ↕ Financials Evidence note News signal Evidence verified

Scroll horizontally for all columns

AI Startup Concentration by City
Bubble size proportional to number of top-40 companies · Hover to explore
Scale:
1
5
10
21

Research Insights

Methodology and sources

Dataset purpose

The China AI Ecosystem Table is a curated intelligence and research dataset. It maps selected firms, public companies, platform providers, research institutions, capital actors, and infrastructure anchors relevant to China's AI ecosystem.

Selection

Inclusion is based on analytical relevance, public evidence availability, and ecosystem role. The dataset is neither a ranking nor a complete census. It includes private firms, listed firms, business units, products, research organizations, and capital providers, which are identified by entity type.

This is a curated ecosystem dataset, not a market ranking, investment recommendation, comprehensive national census, or causal estimate of policy effects. Inclusion reflects analytical relevance and evidence availability. Geographic concentration in this sample does not by itself establish that policy designation caused ecosystem outcomes.

Evidence

Core fields are classified as disclosed, reported, estimated, inferred, or unverified. Primary sources include regulatory filings, prospectuses, stock-exchange documents, government documents, company disclosures, and institutional records. Corroborated reporting from outlets such as Reuters, Bloomberg, Caixin, Financial Times, Nikkei, CNBC, SCMP, TechNode, 36Kr, Pandaily, and specialist sources may be used for private-market or emerging ecosystem information when primary records are unavailable.

Disclosed
Primary source, filing, exchange disclosure, government document, or official institutional statement.
Reported
Credible financial or specialist reporting, shown with source name and date.
Estimated
Third-party estimate, avoided for headline numbers.
Inferred
Analytical interpretation based on documented facts, never displayed as direct fact.
Unverified
Held out of the public fact layer.

Financial data schema

The table presents one concise Financials field per entity so unlike metrics are not mistaken for directly comparable measures. Each displayed figure is classified by type and accompanied by a date or reporting period, claim status, and source.

Financials
A sourced entity-level revenue, financing, transaction, IPO, or dated market-capitalization observation. Where no defensible standalone figure is available, the table states “Not publicly disclosed.”

An intended valuation, fundraising target, press rumor, or parent-company market capitalization is never presented as the funding or valuation of a subsidiary, product, or business unit.

Interpretation

The developmental-orchestration framework is an analytical lens. A domain label identifies a documented or hypothesized ecosystem mechanism. It does not by itself establish that a policy, investor, university, or infrastructure provider caused a company's outcome.

Case design

Beijing is the capstone's primary case. Hangzhou is a contrastive reference. Shanghai, Shenzhen, Hefei, Wuhan, Guangzhou, and national-level entities provide broader ecosystem context. This visualization does not provide a matched comparative design or causal estimate of city policy effects.

Relationship taxonomy

A network edge identifies an exact relationship type. An edge is never drawn from geographic proximity, founder education, general policy relevance, or narrative plausibility alone. Founder education is never represented as a formal spinout or technology-transfer relationship.

  • Financing
  • Formal spinout
  • Founder education
  • Formal university or research partnership
  • Joint laboratory
  • Public investment or guidance-fund investment
  • Public procurement or deployment
  • Cloud, compute, or model dependency
  • Research institution anchor
  • Acquisition

Updates

Research evidence is frozen as of 17 September 2026 for this release. Dynamic news signals are maintained separately and are not evidence for the table's core factual classifications.

Independence and attribution

This independent research visualization does not represent the views or endorsement of Schwarzman College, Tsinghua University, government agencies, companies, investors, advisors, or interview participants. No confidential interview material, private communication, or non-public company information is displayed without explicit authorization.

Link copied