Raising a seed round in AI.
At seed, AI investors want proof that customers depend on the product and that the economics can work once compute bills are real. Here is the bar and who usually writes the checks.
AI / ML
AI investors judge three things before anything else: an advantage the big model providers cannot copy, gross margin after compute, and proof that customers rely on the output, not just try it.
What seed investors expect in AI
Seed in AI is judged on three things before anything else. Show each one on a slide of its own.
Usage that sticks
- Paying customers or pilots turning into contracts
- Retention after the first month, not only sign-ups
- Output quality customers trust enough to act on
A data advantage forming
- Proprietary or hard-to-get data you have rights to use
- Feedback from usage that improves the product
- Evaluations that show you beat a general model on your task
Margins you can explain
- Inference cost per customer or per task
- Gross margin today and how it improves with scale
- A plan if a model provider raises prices or changes terms
How the round comes together
Investor types, not names. Read each guide, then match inside your account.
- 1 · A lead sets the terms
- 2 · Other funds follow
- 3 · Angels fill the rest
Pre-seed & seed funds
Often leads
Most AI seed rounds are led by a seed fund with technical partners or an AI thesis.
Multi-stage VCs
Sometimes
Larger funds write seed checks in AI to secure a place in the Series A.
Angels
Often follows
Operators and researchers fill the round and help with hiring.
Corporate VCs
Sometimes
Cloud, chip and software companies invest for partnerships and compute access.
Accelerators
Now and then
Some teams raise seed straight after a program's demo day.
SAFE or priced round?
Many AI seed rounds start on SAFEs. Larger rounds with a lead often become priced equity with a board seat. If a corporate investor offers compute credits as part of the deal, value them separately from the cash.
The slides investors read first
Traction
Paying users, retention and contract value.
Data moat
What data you own and how usage improves it.
Evals
Your model against general models on your task.
Gross margin
Revenue per customer after inference costs.
The ask
Amount, instrument and the 18 months it buys.
Questions founders ask
How do seed investors judge AI gross margins?
They expect margins below typical software at first, because inference costs real money. What matters is a credible path up: smaller models, caching, routing or better pricing as you grow.
How much do AI startups raise at seed?
It varies with how much compute and research the plan needs. Use the round-size benchmark to see what founders at your stage and in your city actually raised.
Sources
Sources: investors' published criteria for AI companies; cloud and model provider pricing documentation; industry reports. Editorial content, reviewed quarterly. Not financial or legal advice.
Other stages
Related sectors
Who funds this round
Learn the terms
Run the numbers
Raise in your city
Inside Unicorn Factory
Free account
Investor matching is coming to your free account. Create it now and we will match your round the day it opens. Until then, get an honest read on your deck.