What changes when AI becomes part of a product, rather than the whole pitch? These landscapes connect capabilities to the people they serve, the work they do and the businesses built around them.
Large ecosystems, model providers and specialists sit alongside one another. Start with a visual overview, then use the company explorer to compare products, operating boundaries, funding and ownership. Each study adds product contrasts, dated evidence and questions worth investigating. The aim is orientation, not a leaderboard or an exhaustive directory.
Explore the landscapes
Consumer AI Products
36 company groups across learning, creation, companionship and everyday life. What makes a product worth returning to?
Personal AI Agents
23 company groups and a separate watchlist of five cases. Who acts, who approves and what proves completion?
China’s Consumer AI Ecosystem
26 company groups spanning creative tools, learning, personal workflows and conversational hardware. How do these experiences become useful products?
Physical AI
52 companies and platforms spanning models, robot hardware, data, testing and deployment. What turns a demonstration into reliable operation?
How to read these maps
Examples establish a pattern, not its market share. Categories overlap. A company can serve consumers, professionals and businesses through different channels. Its appearances in several studies offer different perspectives, not additional companies to count.
Features are not outcomes. Official product pages describe offers and provider claims, not independently established learning gains, wellbeing, reliability or customer economics. Interpretation and proposed evaluation questions remain separate from those claims.
Funding is a separate signal. Explorers give funding and ownership their own column, apart from product status. Round size, total capital, valuation and secondary share sales are different measures. A corporate product is not a separately funded startup; financing alone does not establish product success.
Read the dates. The latest funding review is dated October 11, 2026. Each disclosure keeps its event date, and product evidence retains its own verification dates. Historical rounds are not guaranteed to be the latest; planned raises and listings are not completed events. Unknown means unverified, not unfunded.
For the technical mechanisms behind these products, see Agents and LLM & VLM.