Australian startups closed just 31 early-stage rounds under $5 million in the three months to 30 June 2026, the lowest quarterly count since Cut Through Venture began tracking the data in 2020, according to its Q2 2026 funding report as reported by Forbes Australia.
The drop came in a quarter when total funding rose. Startups raised $1.7 billion across 64 venture deals and five accelerator rounds, up 60% on a year earlier, but two deals did most of the lifting. Firmus ($725 million) and payments group Airwallex ($460 million) accounted for about 70% of the capital, and the top five deals for about 80%. The result is an ecosystem where the biggest cheques keep getting bigger while the pipeline of first cheques thins out.
AI is now most of the market
AI companies took about 75% of capital and two-thirds of deals in the quarter, up from 64% of both in the March quarter, according to the report. AI featured in 81% of seed rounds.
That concentration is changing the size of early rounds. Median pre-seed and seed raises reached $1.3 million and $4 million, the highest on record. One investor quoted anonymously in the report said many companies that would once have raised $1 million to $2 million were now seeking $5 million to $10 million on an AI narrative, which it said was “boxing early-stage funds like us out of the market.”
Fewer backers at the start
The count of investors writing first cheques is also small by global standards. Side Stage Ventures’ Australian Venture and Startup Report 2026 found 18 local VCs backing seed and pre-seed rounds, compared with more than 500 in Europe and nearly 600 in the US, Forbes Australia reported.
Small rounds have been under pressure for some time. Cut Through’s March 2024 quarter data showed deals under $5 million at a five-year low, with fewer angel investors a major factor, InnovationAus reported at the time.
The case for using the lag
Some local founders and commentators argue Australia trails the US AI market by several years, and that the lag can be useful: the wave of generic AI wrapper startups has already played out in the US, and capital there has moved toward tools built for one industry, one workflow and a buyer already paying people to do the job. On that view, the US deal list works as a forecast for Australian founders building in sectors where the country has depth, such as insurance, freight, mining and agriculture.
The Q2 numbers suggest the constraint is less the ideas than the first cheque.
Jack’s take
I think Australia sits roughly five years behind the US on AI, and that is only an advantage if we act on what the US has already shown us. The gap does not close at Series B, it closes when more founders get their first cheque, and right now that is the part of the market shrinking. I’d add two more things: build for a global market from day one, because Australia is a great first customer and a terrible only one, and keep the people we train here.