China's DeepSeek has resumed its second funding round, The Information reports; separately plans a “significant increase” in the near future in API pricing for AI models, according to website

Private funding rounds at high-profile AI developers have tended to matter less for the round itself than for what the implied valuation and participant list signal about the sector's capital intensity and about state tolerance for the firm's ambitions. The pricing leg is the more market-relevant element: this firm's prior form was to compete on aggressive undercutting of Western model APIs, and a significant price increase, if confirmed on its own channels, would mark a shift from share-grabbing toward monetisation, a sequence seen repeatedly in subsidised challenger markets once scale is established. The distinction worth drawing is between cost recovery driven by compute constraints and genuine pricing power, since the former says little about the incumbent peer set while the latter eases the deflationary pressure cheap Chinese inference has exerted on the listed AI complex. On the funding side, the tells are whether sovereign-linked or strategic capital features and whether the round prices above prior marks, both of which speak to Beijing's backing. Note the sourcing is secondhand reporting plus the company's own website, so confirmation risk is live. Watch for reactions in the Chinese AI-adjacent names and in the compute supply chain that has traded off this firm's efficiency claims.