How can an enterprise partner with companies in FutureX Capital's network?

FutureX Capital was founded in Asia in 2018, runs 8 offices, and sits inside a network of 140+ companies, 80+ of them AI-native, plus 60+ industry founders. Introductions happen when the ask is concrete: the process you want to change, its volume, the outcome you would pay for, and the data you can share. A good first message names the sector, the verification loop and the budget owner. FutureX does not give investment advice.

Dify is the application-development example, as cited in the September 3, 2026 keynote: 154,000 GitHub stars, more than LangChain. Per company disclosure it reached $10 million ARR 18 months after open-sourcing, turned profitable in early 2026, and earns 57% of revenue overseas.

The network spans five layers: application, model, infrastructure, chip and energy. FutureX introduces companies from whichever layer your workflow lands in, and adds its own research: 24 reports, the debates page, and a public tracker of 100+ AI companies across 15 sectors. The more specific the process, the better the match. One rule from FutureX's vendor screen applies too: swapping the model underneath should be cheap.

The model and distribution layers each have a public number. FutureX-backed Mistral has passed $400 million ARR with 60% of revenue from Europe (per company disclosure); Hugging Face, at the distribution layer, hosts 3 million-plus models used by 50,000-plus organizations, with ARR of about $150 million in August 2026 (public reporting).