A pre-seed AI investor needs to believe two things at once: that the technical insight matters, and that it can become a business. A fund's AI portfolio is a starting point, but it does not tell you whether it will write your first cheque, fund your company location or consider the evidence you have today.
This guide compares 16 current first-cheque routes. Some are AI specialists; others back technical companies more broadly. A few use an accelerator or residency. Those differences determine who belongs in your first outreach wave.
Start with the comparison, then use the fit card to narrow it to a short list you can explain. For a broader stage overview, use our pre-seed investor guide. For investors across AI funding stages, see the AI investor directory.
Choose the kind of first cheque you need
- Technical specialist: Useful when architecture, research or a hard engineering constraint is central to the company. Prepare a demonstration and a defensibility argument.
- Business-model partner: Useful when an identifiable buyer and a difficult workflow are the strongest evidence. Explain how AI changes the outcome, not only the interface.
- Programme investor: Useful when the working environment, attendance and company-building support suit the team. Investment terms and programme obligations deserve separate review.
Pre-seed does not always mean pre-revenue or pre-product. Forum's current fund criteria, for example, ask for more readiness than an idea-stage route. AIX also describes product-market-fit signals. Treat the label as an invitation to check the criteria, not as automatic eligibility.
Compare 16 pre-seed AI investors
The list is not a ranking. Each provider has current official evidence of early-stage investing and AI relevance. Where a reviewed source does not state a geographic rule, the table says to confirm it rather than substituting the firm's office location.
| Investor and official criteria | Technical focus | Stage or round role | Geographic filter |
|---|---|---|---|
| 2048 Ventures | AI, deep tech and health/bio | Pre-seed and seed; can lead or follow | US companies |
| ellipsis Ventures | AI core to the product | First cheque through seed; leads and co-invests | Europe first; selected North America and Israel |
| Air Street Capital | AI-first across applications, infrastructure and science | Pre-seed through Series A; usually leads or co-leads | North America and Europe, including the UK |
| AIX Ventures | AI applications, infrastructure and robotics | Pre-seed through Series A; leads and partners | Geographic eligibility not specified in reviewed FAQ |
| Pear | AI applications and infrastructure within a broader technology strategy | PearX pre-seed programme; direct early rounds can be led or co-led | PearX requires a San Francisco Bay Area stay; confirm direct-investment geography |
| Amplify Partners | Technical AI, data infrastructure and developer tools | First company round through Series A; leads | Confirm company eligibility with the investment team |
| Episode 1 | B2B software, AI and technical infrastructure | Pre-seed and seed; prefers to lead, can co-lead or follow | UK and Europe |
| Eniac Ventures | AI infrastructure, vertical AI, data and applications | Pre-seed and seed; current AI seed lead evidence | Confirm geographic eligibility; office location is not a mandate |
| Faber | Applied science, AI, data and computing | Pre-seed and seed; documented AI seed lead | Southern Europe and selected European companies |
| Unusual Ventures | Enterprise AI and technical software | Pre-seed onward; documented AI seed lead | Confirm company geography with the team |
| Crane Venture Partners | AI infrastructure, enterprise technology and technical systems | Inception through seed; documented seed lead | Expanded strategy covers Europe, North America, India and APAC |
| Forum Ventures | B2B software and AI | Distinct accelerator and pre-seed fund; readiness differs | Confirm location eligibility for the specific fund or programme |
| Basis Set | AI-native productivity, infrastructure and automation | Pre-seed through Series A; commonly leads or co-leads early rounds | Confirm company geography; San Francisco base is not an eligibility rule |
| Afore Capital | Technology generalist with current AI-company examples | Pre-seed, including pre-idea and pre-traction | San Francisco residency; confirm attendance fit for founders elsewhere |
| SuperSeed | AI transforming the physical economy | Pre-seed and seed; usually leads, can join | UK company or plan to establish UK headquarters |
| Hoxton Ventures | Sector agnostic, including documented AI companies | Pre-seed and seed; lead or co-lead | Europe, with ambition to build a US market presence |
Find the right first-cheque route
1. 2048 Ventures
Present the proprietary data or technical insight that makes the business difficult to reproduce. Its current company-location filter is the US, even when the founding team has international roots.
Founder route: Official investment team. Review the current criteria.
2. ellipsis Ventures
AI needs to be fundamental, not an added feature. Its criteria distinguish defensibility through data, expertise or demanding computation from a thin application wrapper. Early revenue is not a prerequisite.
Founder route: Public submission form. Review the current criteria.
3. Air Street Capital
A specialist route for companies where AI is central. Its FAQ welcomes early approaches and asks founders to explain the technology, company and round. Show what customer learning exists even before a finished product.
Founder route: Published founder contact email. Review the current criteria.
4. AIX Ventures
AIX includes pre-seed but also describes looking for product-market-fit signals, including revenue. Treat stage and readiness as separate filters. A technically interesting idea may need stronger customer evidence before it matches the published approach.
Founder route: Official investment team. Review the current criteria.
5. Pear
Pear routes pre-seed founders through PearX, a twelve-week accelerator with an in-person Bay Area requirement. That is a programme decision as well as a financing decision. Do not confuse it with Pear's direct early-stage investing.
Founder route: PearX application for pre-seed. Review the current criteria.
6. Amplify Partners
Relevant when the technical architecture is the company rather than a supporting feature. Its first-round approach includes technical founders still turning an insight into a business. Bring an architecture explanation and a clear account of who needs the product.
Founder route: Official investment team. Review the current criteria · Supporting official evidence.
7. Episode 1
Its first-institutional-investor positioning fits a technical B2B team with a clear market problem. Explain the buyer and the advantage, not only the model. Most consumer businesses and conventional hardware sit outside its stated core.
Founder route: Submit Your Deck form. Review the current criteria.
8. Eniac Ventures
Current investor biographies identify pre-seed AI infrastructure and AI-native applications. Match the company to the relevant investor rather than sending the same note to every partner. Show the technical constraint and the customer workflow together.
Founder route: Relevant investor's official team profile. Review the current criteria · Supporting official evidence.
9. Faber
Faber combines a geographic filter with an applied-science thesis. Its AI scope reaches data infrastructure, scientific software and computational biology. Choose the correct pitch route rather than assuming its climate and applied-science strategies share identical criteria.
Founder route: Applied-science pitch route, distinct from climate/ocean route. Review the current criteria · Supporting official evidence.
10. Unusual Ventures
Unusual's current hiring materials identify a pre-seed-to-Series-B investment range, while its company work centres on enterprise AI. The old Spark announcement is not proof that a particular cohort is open today. Use the current investment team.
Founder route: Official investment team; newsletter is not an application. Review the current criteria · Supporting official evidence · Further official evidence.
11. Crane Venture Partners
Crane's September 2026 strategy expands beyond an older Europe-only description. Its technical focus makes architecture and deployment evidence especially useful. Confirm the applicable vehicle and company-location requirements rather than reading eligibility from an office address.
Founder route: Official firm contact and investment team. Review the current criteria · Supporting official evidence · Further official evidence.
12. Forum Ventures
Forum's pre-seed fund expects an MVP plus early revenue or strong design-partner traction. Its FAQ describes prior-funding or revenue indicators and directs earlier founders toward the accelerator. Pick the right route before investing time in a pitch.
Founder route: Public pitch route. Review the current criteria.
13. Basis Set
Its first-capital positioning and AI thesis fit teams building a foundational product, not merely adopting AI internally. The FAQ explicitly includes pre-seed. Use current fund announcements for activity context rather than repeating the FAQ's older fund-total figures.
Founder route: Official investment team. Review the current criteria · Supporting official evidence.
14. Afore Capital
Afore is a generalist first-cheque option, not an AI-exclusive fund. Its residency includes technical and AI founders and makes investment decisions separately from cohort starts. Evaluate the working arrangement as well as the capital; do not assume every dated community page is an open application.
Founder route: Founder-in-residence application. Review the current criteria.
15. SuperSeed
The current thesis is narrower than generic B2B SaaS: AI must improve the physical economy. Manufacturing, construction, logistics and energy can fit. Specialist life-science or materials opportunities require a clear AI angle and an appropriate specialist lead.
Founder route: Public pitch route and published pitch email. Review the current criteria.
16. Hoxton Ventures
Hoxton is a generalist, not a dedicated AI fund. Its early-stage European strategy and US-market ambition are the relevant filters. Make the expansion case concrete without implying that a US customer plan substitutes for the European company mandate.
Founder route: Official Submit Pitch route. Review the current criteria.
What to prepare before approaching an AI investor
Put the technical claim next to the evidence that supports it. If the product improves an existing workflow, show the task, comparison method and result. If the advantage is proprietary data, explain the rights and collection process. If you are creating infrastructure, describe the deployment constraint and the developer who feels it.
Then connect that evidence to a buyer. An unpaid design partner can demonstrate learning, but it is not contracted revenue. A successful benchmark can demonstrate one technical property, but it does not establish customer demand. Label each signal so the investor does not need to reverse-engineer your progress.
- One sentence on the buyer, painful task and intended outcome.
- A short demo or technical result, with its limits stated.
- A defensibility explanation that survives improvements in third-party models.
- Customer conversations, pilot usage or sales, clearly separated.
- The milestone this round funds and the company's capital history.
A worked first-cheque shortlist
Fictional example: A UK team is building AI software for factory maintenance. It has a working prototype, three unpaid design partners and no sales. Its next milestone is a paid deployment that demonstrates reliability on a real production line.
The team would test SuperSeed's physical-economy thesis and UK requirement, Episode 1's technical B2B criteria and AI-specialist routes such as Air Street. It would not mark Forum's fund as a fit solely because the company is called pre-seed; it would compare the actual design-partner evidence with Forum's readiness expectations. A US-only mandate would remain outside the first wave unless the company's location genuinely met it.
The useful output is three or four reasoned conversations, not sixteen identical emails. Put the unknowns beside the strengths: who owns the data, how the system handles failure, and what the paid pilot needs to prove. Use the pre-seed validation sprint to turn those gaps into experiments.
Build your AI first-cheque fit card
Copy one card per investor. Keep the source date beside the claim, and update the next action after every conversation. This makes it easier to request a relevant introduction and to avoid approaching an ineligible firm.
AI FIRST-CHEQUE FIT CARD
Company and country:
AI role: core product / infrastructure / scientific method
Buyer and problem:
Prototype or technical result:
Customer evidence, labelled paid or unpaid:
Data rights and defensibility:
Capital raised to date:
Round amount and milestone:
Investor / relevant partner:
Official mandate URL and date:
Stage / geography / programme fit:
Readiness gap or portfolio conflict:
Public route or consented introduction:
Next action / owner / date:Ask Aurora, Claude or Codex to pressure-test your shortlist
Use your own company facts and the linked official sources. Review the output before acting; the prompt is a research aid, not an automatic outreach tool.
Act as a critical investor-research assistant. I am preparing a pre-seed AI raise.
Use only the official source URLs and company facts I provide. Do not invent eligibility, introductions or investor interest.
For each firm, distinguish AI specialist from generalist, direct investment from accelerator, and office location from investment eligibility.
Compare technical evidence, buyer evidence, geography, previous funding and programme attendance requirements.
Return a table: firm, source-backed fit, unresolved question, reason to exclude, next action.
Treat unpaid pilots as unpaid and mark unknown requirements as unknown.
Draft one concise introduction request for my review; do not send anything.Who are the top pre-seed investors?
The most useful investor is one that fits your company, not the largest name in a search result. For an AI company, compare technical focus, stage, geographic eligibility, readiness, relevant partner and the kind of first cheque you want. Air Street and Basis Set offer AI-focused approaches; firms such as Afore and Hoxton are broader early-stage options. Their fit depends on the company and route, not a universal league table.
Turn your research into a working raise
Organize the shortlist in Finta, keep the evidence and relationship context beside each contact, and track the next step with Aurora. Start with a few well-supported targets and expand as you learn. Our comprehensive pre-seed guide connects that research to the wider process.
Methodology and research date
Finta Editorial Team reviewed official criteria, team biographies and dated investment announcements on November 12, 2026. Firms and rebrands are counted once. AI specialists and generalists are identified separately; residency, accelerator and direct-investment routes are not interchangeable. An example investment establishes activity, not an open allocation or acceptance promise. Verify the current mandate and any terms directly. General information, not investment advice.