Recommended Articles
Fundraising

Why "Pre-Seed" Now Means $30M Checks in Robotics and Embodied AI

By Angeled · October 23, 2026 · Fundraising
Why "Pre-Seed" Now Means $30M Checks in Robotics and Embodied AI
Why "Pre-Seed" Now Means $30M Checks in Robotics and Embodied AI

The take

The label "pre-seed" no longer means small checks. In October 2026, Enact Intelligence — a Zurich-based robotics startup — emerged from stealth with a $30 million pre-seed round led by redalpine, Notion Capital, and firstminute, joined by angels including Mistral's Arthur Mensch, Yann LeCun, and Hugging Face co-founder Thomas Wolf. Sifted called it the largest pre-seed in Switzerland's history. Meanwhile, Nirva, an AI-pendant hardware startup founded by former Meta Reality Labs staff, raised a $25 million pre-Series A, bringing its total funding to $33 million. Hardware-heavy robotics and embodied AI now price like late-stage software rounds at birth — and founders and investors need to update their mental models of what "early" means.

Background and context

In software venture capital, "pre-seed" traditionally meant a few hundred thousand to a couple of million dollars raised to test an idea before product-market fit. The label reflected a stage: pre-product, pre-revenue, pre-traction. Over the past decade, round labels have inflated steadily — Series A rounds became what Series B used to be — but the pre-seed tier held roughly steady because software's earliest milestones were cheap to reach: a prototype, some user interviews, maybe early revenue.

Embodied AI breaks that logic. Building a robot is not like building an app. Before a robotics startup can demonstrate anything meaningful, it must design and fabricate physical hardware: actuators, sensors, power systems, chassis, and the manufacturing pipeline that produces them at quality. Each of those components has lead times, tooling costs, and iteration cycles measured in months and real dollars. A robotics company that has not yet shipped cannot credibly demonstrate traction with a slide deck; it needs functioning machines, which means it needs millions before it can claim the "traction" that software investors expect from a Series A.

What the data shows

Enact Intelligence's $30 million pre-seed is instructive in both size and composition. The investor list — three institutional venture firms plus three of the most recognizable names in AI research — signals that this round was not priced on traction but on conviction: conviction in the team, the technical approach, and the thesis that embodied AI is the next platform shift. When Yann LeCun and the CEO of Mistral write angel checks into a pre-seed, the round is being underwritten on scientific credibility, not revenue curves.

Nirva's $25 million pre-Series A tells a parallel story in consumer hardware. Founded by ex-Meta Reality Labs staff, the team brings institutional knowledge of exactly how expensive and difficult consumer AI hardware is to build — and investors priced that pedigree accordingly. At $33 million in total funding before a public launch, Nirva is capitalized like a growth-stage software company while still being, technically, an early-stage hardware bet. The through-line: capital intensity is being priced into the round label itself. "Pre-seed" now describes a company's position on the product timeline, not its capitalization.

What it means for founders

For founders raising early in hardware, robotics, or embodied AI, the price of entry has risen — but so has the ceiling. The good news is that investors clearly understand capital intensity and will write large early checks for the right teams. The bad news is that "the right teams" is now a narrower filter than it used to be: institutional-grade technical pedigree, demonstrated hardware execution, and often a prior working relationship between founders and investors are becoming the de facto requirements to unlock these supersized early rounds.

If you are a first-time founder without that pedigree, the playbook is unforgiving but clear: build the most impressive physical demonstration possible on the smallest possible budget — a single working robot, one compelling hardware prototype — and use it to recruit the technical credibility you lack, whether as a co-founder, advisor, or early hire. Investors at this stage are underwriting the team's ability to survive the hardware iteration cycle; everything in your pitch should demonstrate that you understand actuators, sensing, supply chains, and manufacturing timelines, not just models and software. And do not anchor your raise size to the old pre-seed norms — under-raising in hardware is fatal, because running out of money mid-iteration leaves you with nothing demonstrable.

What it means for investors

For investors, the supersized pre-seed changes the underwriting model. Pre-seed used to be optionality capital: many small bets, expecting most to fail, hoping one breaks out. A $30 million pre-seed cannot be an optionality bet — it is a conviction position, and it demands the diligence posture of a Series B: deep technical review, reference checks on the team's hardware track record, and a thesis-level commitment to the category.

This has portfolio-construction consequences. Funds that historically wrote fifty $500K pre-seed checks must now decide whether to write five $5M checks or concentrate further into two or three $15M+ positions. The risk profile also shifts: hardware startups fail differently than software startups — they fail expensively, mid-iteration, when a supplier delays or a design flaw requires a rebuild. Investors need reserves calibrated to hardware timelines, not software timelines. The funds leading these rounds — redalpine, Notion Capital, firstminute — are effectively signaling that pre-seed in embodied AI is now a specialist's game. Generalist spray-and-pray does not survive contact with a bill of materials.

Angeled.com connects founders with investors.

Sifted reporting, October 22, 2026, on Enact Intelligence's $30M pre-seed (redalpine, Notion Capital, firstminute; angels Arthur Mensch, Yann LeCun, Thomas Wolf), described as Switzerland's largest pre-seed.
Company and press reporting on Nirva's $25M pre-Series A ($33M total raised), founded by ex-Meta Reality Labs staff.

Read Next