Climate tech venture funding just had its strongest first half since 2022. According to investment tracker Currence, startups pulled in $26.1 billion in venture funding during the first six months of 2026, up 55 percent year-over-year, as billions chased the infrastructure underneath the AI boom.

Here is the twist: investors were not backing “climate” broadly. They were backing the compute power chain. Low-carbon datacenter developers alone accounted for 34 percent of all climate venture funding, up from just 3 percent a year earlier. DayOne’s $4.5 billion and Nscale’s $2 billion Series C rounds made up roughly a quarter of all investment between them. In other words, the biggest winner in climate venture is the part that feeds AI hunger.

That concentration has already changed how the industry draws the boundaries of climate tech. Currence effectively says the space is bigger than it was six years ago, with datacenters driving themes that used to live in separate buckets. The report argues datacenter development connects to long-duration energy storage, advanced nuclear, geothermal power, robotics, and even space-based solar. The logic is straightforward: if you need more compute, you need more “speed-to-power,” and investors increasingly treat power sourcing as a core product feature rather than an afterthought.

Currence’s report is also explicit about classification. It counts datacenter developers that rely primarily on clean power or make sustainability central to their business, while excluding conventional facilities and chipmakers. That distinction matters because it reshapes where capital is allowed to flow on paper, and paper is how investors and founders signal where the market is going. If your pitch sounds like a regular data center, you may not fit the definition that is attracting AI-adjacent climate dollars. If you can show a credible “route to power,” you’re more likely to land inside the currently investable version of climate tech.

The shift is showing up in how money is timed and sized. Currence says investors are writing large early stage checks to nuclear startups years before they are expected to generate electricity. The bet is that AI's long-term appetite for power will eventually justify today’s valuations. That is a very different timeline than traditional energy project finance, where delays and permitting risk can dominate the investment story. Here, venture capital is effectively acting like a bridge fund for power supply that might only fully materialize later.

The ripple effects go beyond generation. Earth observation funding has tripled as developers seek larger stores of real-world data to train AI models. Robotics startups building foundational models, training data, and simulation platforms raised nearly four times as much as the next biggest innovation category. These categories look like “AI progress,” but Currence is tying them back to the same engine: the compute buildout needs power, and the power buildout needs execution at scale.

At the same time, some climate themes are losing oxygen. Carbon-related equity funding fell 61 percent to its weakest first half since 2020. Deal count dropped 25 percent even as funding surged, and the ten largest rounds accounted for 42 percent of all investment. This is the classic pattern of a money rush that compresses opportunity: fewer deals get funded, and the biggest rounds get bigger, especially when they sit in the direct path of the AI infrastructure supply chain.

Currence also frames the overall move as a convergence of venture and infrastructure finance. The report says the concentration of capital in huge datacenter and energy projects is pushing climate venture funding closer to infrastructure finance. And if you zoom out, that is the key strategic stake for executives, founders, and board members: climate venture is not just growing. It is re-sorting. Capital is following AI infrastructure: datacenters, power generation, grid capacity, and whatever else keeps the compute boom switched on. The companies that will keep getting attention are the ones that can credibly connect their product to that “speed-to-power” reality, not just to the broad idea of decarbonization.