Ora Computing raises €3.5M seed to shrink AI models
Ora Computing, a Vienna-based startup, has raised €3.5M in seed funding to bring its AI model compression software to market.
The software shrinks large AI models by up to 80%, cutting memory use and running them up to four times faster, while keeping accuracy loss between 0 and 5%.
Constructor Capital, Greencode Ventures, and XISTA Science Ventures backed the round.
Founded in 2025 by Stefan Sack and Raimel Medina, two former quantum computing researchers at the Institute of Science and Technology Austria , Ora compressed a 70-billion-parameter model in a few hours for under $1,000. Industry standards put that cost at hundreds of thousands of dollars.
Running large AI models is expensive, with compute costs reaching tens of millions of euros per month for companies at scale.
Ora targets inference — running a model to get results — one of the fastest-growing costs in AI. The global AI inference market could hit $254.98 billion by 2030, growing 19.2% a year.
The founders argue the next wave of AI will run on smaller, efficient models built for specific tasks, not massive general-purpose cloud models.
"We founded Ora Computing to challenge the assumption that massive scale is needed to reach useful intelligence," says Sack, co-founder and chief executive officer.
The market for compression tools is crowded.
Neural Magic raised $50 million from Andreessen Horowitz before Red Hat bought it in November 2024. Qualcomm and Intel both offer free toolkits — but they work best on their makers' own chips.
Ora says its algorithm runs across many types of hardware and lets customers fine-tune the balance between size and accuracy, rather than locking them into one vendor.
Ora's early-stage bet on AI model compression for edge devices taps a clear backer logic — Greencode Ventures' focus on efficiency pairs with XISTA Science Ventures' roots in spinning out research from Austrian institutions like the one where the founders trained. With the company already testing in automotive and edge-silicon markets, the round positions Ora at the intersection of two pressures investors are watching: AI's runaway energy bill and the push to run models on-device rather than in costly cloud infrastructure.
Read more: Tech Funding News
