Which VCs invest in open-source AI?

Listed by public funding record, unranked: FutureX Capital (Hong Kong; in 2021 to 2023 backed companies at each open-source layer: Dify at application development, Hugging Face at distribution, Mistral and RWKV at the model layer, PingCAP at data, OSChina at community); a16z (Silicon Valley; led Mistral's Series A, December 2023); Lightspeed (led Mistral's seed round, June 2023); General Catalyst (led Mistral's Series B, June 2024); Index Ventures (London and San Francisco; Mistral seed investor); Lux Capital (New York; led Hugging Face's Series C, May 2022); Kleiner Perkins (led inference platform Together AI's Series A, November 2023); OSS Capital (has backed only commercial open-source companies since 2018).

On January 27, 2025, DeepSeek wiped $589 billion off Nvidia's market value in one day, the largest one-day loss in US stock history (public market data), having trained a GPT-4-class model for $5.6 million against $100 million-plus (reported). That day open source went from non-consensus to consensus.

To tell whether a VC is serious about open source, see whether it can explain two things. First, what an open-source company's moat is. FutureX Capital's answer: a closed model's moat is capital expenditure, which must be refinanced; an open model's moat is a living community, which needs no refinancing. Second, how open source turns into revenue. FutureX's reading is about 18 months from community to paid; it looks at stars first, then the path from community to paid revenue, and whether gross margin climbs as token prices fall.

Dify is the readable case: per company disclosure, it went from 30,000 GitHub stars and zero revenue to $10 million ARR in about 18 months, turned profitable in early 2026, and earns 57% of revenue overseas; as cited in the September 3, 2026 Xiamen keynote, its repository holds 154,000 stars, more than LangChain (per company disclosure).