How to Find Investors for an AI Startup: Stage Mix, States and Fund Sizes

By Muhammad Tayyab Shabbir. Last updated 10 Nov 2026. Data as of November 2026. Markdown version

To find investors for an AI startup, start from our list of 1,576 AI investors and narrow it by stage, place and fund size. Seed is the stage they name most (1,082 of the 1,370 with a stage on file), ahead of pre-seed. 71% are in the US, where California leads with 464. The median fund on file from SEC Form D is $60M, above the $50M for every investor we hold with such a figure.

By type, the 1,576 are 1,413 VC firms, 33 corporate venture arms, 30 angel networks, 27 accelerators and 24 SEIS or EIS funds, with 49 of other types. The tags these investors most often carry beside AI are SaaS and B2B software (1,189), healthtech and digital health (651) and fintech (614).

What do AI investors look for?

An AI pitch is read for what the company owns that a competitor with the same models could not copy. Investors ask where the data comes from, what the product costs to run for each customer and whether usage holds up after the first trial.

• What is yours and what is rented? Models, data and distribution can each be built, licensed or borrowed, and the answer shapes the margin.
• What does it cost to serve one customer? Compute is a variable cost, so the gross margin of an AI product has to be shown, not assumed.
• Are customers still using it after the pilot? Retention after the first months separates a working process from an experiment.

Which stage do AI investors back?

Among the 1,370 of the 1,576 with a stage on file (a firm can state several), seed comes first with 1,082 (79%), then pre-seed (633), Series A (500), growth (295) and Series B (204). Set against every investor we hold, the group leans toward pre-seed (46% here, 39% overall) and away from growth (22% here, 28% overall). With about four fifths naming seed, a seed round has the widest field on this list.

Where are AI investors based?

71% of the list is American: 1,123 of the 1,576 are in the US, 214 in the UK and 230 in the rest of Europe. Among the 1,099 US investors with a state on file, California holds about two fifths (464), more than its 36% share of all US investors we hold. New York is a distant second with 159, then Massachusetts (61) and Florida (60).

By city, counting the head offices of every type of investor, the largest groups are San Francisco (150), New York (131), Palo Alto (64) and Menlo Park (47).

Metro area lists for this sector, each with its own row count: AI investors in the San Francisco Bay Area (364), AI investors in Silicon Valley (172), AI investors in Los Angeles (63), AI investors in Miami and South Florida (39), AI investors in Austin (30), AI investors in Chicago (28).

The same investors as lists by region: AI venture capital firms in the US (1,112), AI investors in the UK (214), AI investors in Europe (230).

How big are AI funds?

Across the 867 of the 1,576 AI investors with a fund size from SEC Form D on file, the median is $60M, above the $50M for every investor we hold with such a figure. About a quarter of these funds are under $25M and about two fifths are $100M or more. The figure is the fund as filed, not a check size, and a firm's main fund on our record, which is not always its newest; for 180 of the 867 the offering amount is indefinite, so it is the amount sold so far.

Which AI investors can you look at first?

The AI tag is broad, so on its own it is a weak filter. Look at what the firm backed before the current wave, and at whether it funds models, tools for builders or applications. Then keep the ones that invest at your layer.

The 5 firms below are among the free rows of our list of AI venture capital firms in the US (1,112): each names the sector in its own description, and they stand in that list's order, by our activity score, which is no measure of quality.

• Kleiner Perkins

  HQ: Menlo Park, California · Type: VC firm · Stages (our research): Pre-seed, Seed, Series A, Series B, Growth · Latest activity: Form D filing, Aug 2026 · Fund size on file: $995M (SEC Form D, 2026)

  Venture capital firm headquartered in Menlo Park, CA, with a portfolio spanning established technology names and newer AI companies.

• Toyota Ventures

  HQ: Los Altos, California · Type: Corporate VC · Stages (our research): Seed, Series A · Latest activity: Portfolio news, Oct 2026 (our research) · Private fund assets: $928M (SEC Form ADV)

  Los Altos early stage venture arm of Toyota investing in frontier tech and climate tech, including AI, robotics, computing, aerospace, mobility, energy storage and carbon capture.

• Craft Ventures

  HQ: San Francisco, California · Type: VC firm · Stages (our research): Seed, Series A, Growth · Latest activity: Form D filing, Sep 2026 · Fund size on file: $712M (SEC Form D, 2022)

  San Francisco venture firm run by senior operators, backing B2B software, AI and other technology companies at several stages.

• Madrona

  HQ: Seattle, Washington · Type: VC firm · Stages (our research): Pre-seed, Seed, Series A, Series B, Growth · Latest activity: New deal, Oct 2026 (our research) · Fund size on file: $490M (SEC Form D, 2025)

  Seattle venture firm backing technology founders at seed, early and acceleration stages in Seattle, the Bay Area and beyond, across AI, biotech, consumer, infrastructure, deep tech and security.

• Safar Partners

  HQ: Cambridge, Massachusetts · Type: VC firm · Stages (our research): Seed · Latest activity: New deal, Aug 2026 (our research) · Fund size on file: $300M (SEC Form D, 2023)

  Cambridge, MA technology venture fund investing in startups spun out of universities, mainly MIT, Harvard and the University of Rochester. Covers cleantech and advanced materials, AI and robotics, and life sciences.

Which US AI investors filed with the SEC most recently?

512 of the 1,123 US AI investors (46%) have a Form D filing dated in the 12 months to 6 Nov 2026, a first notice or the yearly amendment of an older fund, above the 38% for all US investors we hold. Our file of new fund notices for the 8 weeks to 6 Nov 2026 holds 107 from 76 of these investors, among them Quiet Capital, LightX Capital and Steel Atlas.

What should an AI founder have ready before the first email?

An AI company is asked what it costs to run and what it owns, so bring both answers.

• Gross margin after compute, per customer or per task.
• Where your training and test data comes from, and the rights you hold to it.
• How many of the customers who started a trial are still using the product three months later.

Words AI investors use

Inference cost

What it costs in compute each time the model produces an answer.

Fine-tuning

Further training of an existing model on your own data.

Evaluation set

A fixed set of test cases used to measure whether the model is accurate enough.

Data rights

The legal permission to train on, store and use a set of data.

A common slip. Describing the model before the problem. Say who uses the product and what it replaces, then how it works.

How do you reach AI investors?

93% of the list has a named decision-maker on file (1,473 of 1,576, 7,399 names in all), above the 84% across the whole database. About two fifths have a firm email address (631), and 426 publish how they want to be pitched, such as a form or an application page.

Use a pitch route where a firm publishes one, and otherwise write to one named person at a time, starting from the free outreach email templates; the full list is AI investors, all 1,576 investors, one per row, the first 25 free to read, and the matching tool opens with AI already picked.

This list has 1,576 rows and a single-list file covers up to 300, so its rows are sold inside the Global Bundle file.

Questions