--- url: 'https://qubit.capital/blog/top-investors-backing-ai-startups' title: 'Top AI Investors 2026: Check Sizes & Active Portfolios' author: name: Vaibhav Totuka url: 'https://qubit.capital/blog/author/vaibhav-totuka' date: '2026-05-11T13:05:00+05:30' modified: '2026-07-24T19:28:24+05:30' type: post categories: - Industry-Specific Insights image: 'https://qubit.capital/wp-content/uploads/2026/05/top-investors-backing-ai-startups.webp' published: true --- # Top AI Investors 2026: Check Sizes & Active Portfolios Nearly every VC has bolted an AI thesis onto its homepage, but few have closed a dedicated AI fund this year. That’s the gap you’re facing raising seed through Series B for foundation models, AI infrastructure, or applied AI. You’ve read two or three lists of top investors backing AI startups, each a logo wall with no read on check size. This list of top investors backing AI startups pairs each with your sub-sector, stage, check size, and speed signal to term sheet. If your runway is under nine months, start with the firms ranked first. Larger, slower names further down belong on a parallel track, not your first calls. How we ranked this list A firm needs at least one AI round in our freshness window, a check size overlapping the $2M to $40M range most readers raise, and a verifiable primary source for every figure. The list skips growth-stage and Series C+ AI investors and generalist funds with no demonstrated AI check history. The bigger risk is not picking the wrong firm but approaching the right ones out of order for your stage, a read that comes from our ongoing work advising AI founders. Table of Contents - [The 12 Top Investors Backing AI Startups](#the-12-top-investors-backing-ai-startups) - [Why the Obvious Answer Is Wrong](#why-the-obvious-answer-is-wrong) [How to Spot a Real AI Investor Versus a Trend-Chaser](#how-to-spot-a-real-ai-investor-versus-a-trend-chaser) - [Know Your AI Sub-Sector Before You Build the List](#know-your-ai-sub-sector-before-you-build-the-list) - [Conclusion](#conclusion) ## The 12 Top Investors Backing AI Startups ![Coatue Management logo](data:image/png;base64,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)1. Coatue ManagementFund Size VintageApproximately $70B AUMTypical checkGrowth-stage and crossover, not seedBest forSeries C+ AI companies scaling fastLimitationNot structured for seed or Series A checks Coatue Management fits growth-stage AI companies, not [founders raising a first institutional round](https://qubit.capital/blog/how-to-raise-money-for-ai-startup). The firm runs a hedge-fund-style strategy that crosses over between public markets and late-stage private bets. It closed 39 private deals in 2025, per [Coatue Joins $2.3 Billion Bet on Hot AI Coding Platform](https://www.institutionalinvestor.com/article/coatue-joins-23-billion-bet-hot-ai-coding-platform). One of those was co-leading Anysphere’s Cursor round, a $2.3B Series D with Accel at a $29.3B valuation. In 2026 Coatue launched an AI crossover fund, per [Laffont’s Coatue Plans New AI, Tech Fund as Firm Shifts Long-Only Strategy](https://www.bloomberg.com/news/articles/2026-03-20/laffont-s-70-billion-coatue-looks-to-launch-new-ai-tech-fund). The vehicle can hold public stock, private growth stakes, or cash, signaling Coatue isn’t built for seed-stage checks. Founder Philippe Laffont is steering that shift, per [Philippe Laffont’s $70bn Coatue pivots to launch AI crossover fund](https://pe-insights.com/philippe-laffonts-70bn-coatue-pivots-to-launch-ai-crossover-fund/). He expects AI companies to reach $1.9T in combined revenue within five to ten years. Its footprint backs that up: per [Coatue – 2026 Investor Profile, Portfolio, Team & Exits](https://tracxn.com/d/private-equity/coatue/__neb04ukttPFlKk9SEuR86jzhUs0jtf8EHm92gQaD8r8), Coatue holds 319 companies across 319 funding rounds. What we seeCoatue’s diligence runs through a public-markets lens, grilling growth-stage companies on unit economics and compute costs. [Term sheets can move fast once the firm commits](https://qubit.capital/blog/negotiating-ai-startup-valuation-equity), closer to hedge-fund speed than typical VC pacing. [Founders raising a seed or Series A round](https://qubit.capital/blog/funding-types-for-ai-startups) are usually a poor fit here. ![Insight Partners logo](data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAEAAAABACAMAAACdt4HsAAAAUVBMVEVHcEwWKDQXKTQYKjUXKTQXKTQXKTQXKTQXKTQXKTT/UxoAJzb///8RKDQwMDcAEiLcTB/8UhrGRyEHHyvZ3N1MMzNSXWSytbh+houPlZlweH3BAGwhAAAACnRSTlMAH+v/wIanYNw8+ZKhQQAAAgBJREFUWIWlV+2SgyAM1KJiq0jvQLB9/wc9omcLkigM+6vTYdd8EZKqolDztmt6h6ZreU0eo8gNO6DJELlH7H+NexKd9zgd0PPrr5/QV4kLK7pzOqA7+/ztms/YjTSCp9ABRCTaVD5jbSEfVcjiIwrJ/u84xOGey2cszIWfvyGEBCgljgI3n+/Xj/gNYbXW1i5GSRUqdIQDYp7n6flY4X7OxmGx4zhac5D4OhHWvxBi2AR+pkEIZ71SUr6dhDbSP9jTGfgIfD2XTDuJJVDgqAGEABOTExjfvhc9mUJUQL5Agfnp2KKA9B9MwDFBwPpONMCvYz4hICEKOshETRQxKsAkJHMM/uO4B5TAGgTjmwA+IHxCQC2RAMNDQMRAxRa4IKD3ODkGLghoI0nOgmssaCMnXIjqAK4k+oyhAlsMQw9cGtCXCC/l+DbBdUgVUGsIXwf+tcAG1xCMjvpBgsA8rTDmBfRFqehkfxrEh95h7XuScWOFIJ6l8aHkBxibQRrTCwlFm1PKGHjGZUJR51xnDFkNBUGT1dIQ8JymiqHOaesImpyHBcM952lDsL+uSY8rhs+Yc/q8XxtwjMJhwJhIAW9K8q/kEE44z3kg+MHQfDJkUd8PhqzyMa940CwfdcuH7fJxv3zhKF95quKlqypf+6rixXO1omz1BRQu37tI0vr/B2DzRAyGF80UAAAAAElFTkSuQmCC)2. Insight PartnersFund Size Vintage$12.5BTypical check$5M to $500M+ per company (flagship funds, firm-wide)Best forApplied AI startups building on proprietary data, seed through growthLimitationCheck range isn’t AI-specific or stage-banded; Fund XIII missed its $20B target Insight Partners is a growth-stage software investor that also writes checks into applied AI companies. It closed $12.5B across [Fund XIII, a buyout co-invest fund, and Opportunities Fund II](https://www.insightpartners.com/ideas/global-software-investor-insight-partners-closes-on-12-5b-in-capital-to-invest-in-the-next-generation-of-software-leaders/) in January 2025, pushing regulatory AUM past $90B. The firm has put roughly $4B into AI companies over five years, adding 64 new investments in 2025, per its [2025 year-in-review](https://www.insightpartners.com/ideas/insight-at-30-our-2025-year-in-review/). Fund XIII closed at $12.5B against an original $20B target, a signal on Insight’s deployment pace, according to [Venture Capital Journal](https://www.venturecapitaljournal.com/unlucky-13-for-insight-as-fundraising-falls-short/). George Mathew, who leads Insight’s AI investing, told [Crunchbase News](https://news.crunchbase.com/ai-robotics/insight-partners-investment-portfolio-george-mathew/) he backs “companies that are building domain-specific models on top of private data sets with great user experiences and workflows.” Insight led [Pasito’s $21M Series A](https://www.insightpartners.com/ideas/pasito-raises-21-million-series-a-led-by-insight-partners-to-build-ai-workspace-for-the-benefits-industry/) in February 2026, a live read on its applied-AI check size. What we seeExpect diligence to press on data moats and workflow depth, not just model performance. Insight’s AI team comes from operating roles, not just finance, so questions on differentiation tend to go deep. A firm this size can move slowly on smaller applied-AI checks relative to its growth-stage funds. ![Thrive Capital logo](data:image/png;base64,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)3. Thrive CapitalFund Size Vintage$10B totalTypical checkUndisclosed, drawn from a $1B early-stage poolBest forFounders with high-conviction, breakout momentum storiesLimitationOnly about a dozen new bets a year firmwide Thrive Capital fits founders who want a concentrated backer willing to write large checks over many years, not a spray-and-pray fund. The firm closed a [$10B fund in February 2026](https://pitchbook.com/news/articles/thrive-raises-10b-as-its-lps-anticipate-mega-ipo-windfall), earmarking $1B of it for early-stage deals and $9B for late-stage follow-ons. According to [Bloomberg](https://www.bloomberg.com/news/articles/2026-02-17/josh-kushner-s-thrive-capital-raises-10-billion-in-new-funding), founder Josh Kushner runs the firm on a concentrated, flexible “no-rules” model. That means backing only about a dozen new companies a year, spread across every stage from seed to growth. In December 2025 Thrive put roughly $1B into OpenAI at a [$285B valuation](https://www.cnbc.com/2026/02/25/thrive-capital-openai-joshua-kushner.html), a bet Kushner has pushed since a $29B entry. It also led Anysphere’s (Cursor) [$900M Series C](https://news.crunchbase.com/ai/anysphere-cursor-venture-funding-thrive/) at a $9.9B valuation, having backed the coding startup since an earlier $100M round. What we seeThrive’s OpenAI relationship runs both directions. The firm backs OpenAI, and OpenAI has taken a stake in [Thrive Holdings](https://techcrunch.com/2025/12/01/openais-investment-into-thrive-holdings-is-its-latest-circular-deal/), Thrive’s own AI rollup vehicle. Its [SpaceX and Cursor stakes](https://www.forbes.com/sites/aliciapark/2026/04/24/billionaire-josh-kushners-venture-firm-invests-in-san-francisco-giants-its-first-sports-bet/) alone top $10B, so attention concentrates on very few names. ![Lightspeed Venture Partners logo](data:image/png;base64,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)4. Lightspeed Venture PartnersFund Size VintageOver $9BTypical checkSeed through Series B stage roundsBest forFounders who want operator-turned-investor partnersLimitationGeneralist multi-sector fund, not an AI-only vehicle Lightspeed Venture Partners fits founders who want proof of AI capital already deployed, not just AI branding on a website. The firm has backed 165 AI-native companies and deployed more than $5.5 billion into them, mostly at Seed through Series B. That count comes from [Lightspeed Closes Over $9B in New Funds Aimed at Backing Category Definers Worldwide](https://www.businesswire.com/news/home/20251215319417/en/Lightspeed-Closes-Over-$9B-in-New-Funds-Aimed-at-Backing-Category-Definers-Worldwide). That raise pushed Lightspeed’s global assets under management past $40 billion, per [Lightspeed raises record $9B in fresh capital](https://techcrunch.com/2025/12/15/lightspeed-raises-record-9b-in-fresh-capital/). Partner Guru Chahal co-founded Avi Networks before VMware acquired it. He now backs enterprise AI companies like Exa, per [Lightspeed Team: Guru Chahal, Partner. US. Enterprise.](https://lsvp.com/team-member/guru-chahal/). Only one outside list ranks Lightspeed among AI investors, but its operator-heavy partner bench is the real signal. What we seePartners with founder backgrounds like Chahal’s tend to push harder on product differentiation than generalist check-writers. Expect diligence to probe compute costs and defensibility, not just growth metrics. That bar can slow a process down if the model story is thin. ![Founders Fund logo](data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAEAAAABACAYAAACqaXHeAAAAlUlEQVR42u3asQ2DQBBFwTvk0AF0QIDkClywAzoiogISkHBmAs6F/HkljLTBrrbu49RKcF0JDwAAAAAAAAAAAAAAAAAie/yGPhqgzuviHgAAAAAAAAAAAAAAAAAAANLuAXdr2QCf7TACAAAAAAAAAAAAAAAAAAAgbR2+zm80QH2+3v4DAAAAAAAAAAAAAAAAAAAAgKz+QrAQ2eGfHvAAAAAASUVORK5CYII=)5. Founders FundFund Size Vintage$6B in capital commitmentsTypical check$1B+ per company, growth stageBest forGrowth-stage AI leaders, not first checksLimitationChecks run far above seed-to-Series-B range Founders Fund is nearing $6 billion for its growth fund, per [Founders Fund nears $6B close for latest growth fund, sources say](https://techcrunch.com/2026/03/09/founders-fund-nears-6-billion-close-for-latest-growth-fund-sources-say/). $1.25B went to Anthropic, per [Founders Fund raises $6B after burning through $4.6B in under a year on Anthropic, Anduril, and OpenAI](https://thenextweb.com/news/founders-fund-6-billion-ai-venture). The predecessor $4.6 billion fund went out in under a year, spread across roughly a dozen companies. Partner Trae Stephens leads its defense and AI bets, per [Anduril’s Trae Stephens says defense tech is headed for a shakeout](https://fortune.com/2026/06/05/anduril-trae-stephens-founders-fund-defense-tech-ipo-palantir-venture-capital/). He was an early Palantir employee running defense and intelligence teams before co-founding Anduril in 2017. Its checks run far larger than a typical seed-to-Series-B raise from an AI startup. What we seeFounders Fund moves in size, not volume. A dozen portfolio bets a year means most founders raising a seed or Series B round won’t get a meeting, let alone a term sheet. ![Redpoint Ventures logo](data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAEAAAABACAMAAACdt4HsAAAAGFBMVEVHcEzhEh3iEh3hEh3kEh3jEh3gEh3dEh1tI2dBAAAAB3RSTlMAyECbH2/owaOxAgAAALNJREFUWIXtkkkSwyAMBK0F9P8fJ3ZsQRmMRDiloj7TBTPMtgVBEPwtjGs+Ciz5SURWfHr7spABdl9oTir3cT58yVM+aWSUC57wQY8n9SW59ePRdL6k4P5I/pw/X1Lh9K/QqPXNZdDQhHLDlaGEznfflQFaq8IeY+fWGmuMPNbNMTaltQzHmGx/+JFk68MM4/rtDEb9VgazfqU/Rkf9yrf1K50xuupX2jEmmOPpG4IgCH6YF80dETarYX5NAAAAAElFTkSuQmCC)6. Redpoint VenturesFund Size Vintage$650MTypical checkSeed to Series B leadsBest forApplied and vertical AI, not just foundation modelsLimitationFund size held flat versus 2022, no step-up in scale Redpoint Ventures fits founders building applied or vertical AI products, not pure foundation-model plays. The firm’s tenth early-stage fund closed at $650M in 2025, the same size as the one before it, per [TechCrunch’s report](https://techcrunch.com/2025/05/15/redpoint-raises-650m-3-years-after-its-last-big-early-stage-fund/). Twelve named AI deals in 2025 skewed applied and vertical, not foundation models, per Redpoint’s [2026 outlook post](https://www.redpoint.com/content-hub/written/2026-outlook/). Managing director Alex Bard led the Series A into procurement startup Levelpath and took a board seat, per [coverage of the round](https://www.prnewswire.com/news-releases/levelpath-raised-44-5m-from-a-benchmark-led-seed-round-followed-by-a-redpoint-ventures-led-series-a-and-will-now-launch-the-first-ai-powered-mobile-first-procurement-platform-that-makes-the-procurement-experience-delightful-301929788.html). Redpoint’s early-stage practice has four managing partners, including Erica Brescia, GitHub’s former chief operating officer. Portfolio company Poolside, an AI coding startup founded by ex-Redpoint partner Jason Warner, backs the firm’s alumni, per [Redpoint’s page on Poolside](https://www.redpoint.com/companies/poolside/). What we seeRedpoint’s early-stage checks come from full-time investors, not associates, so expect the partner sourcing the deal to also sit on the board. Diligence conversations with this team tend to probe how a startup’s product differs once the underlying model commoditizes, not just current traction. ![Radical Ventures logo](data:image/png;base64,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)7. Radical VenturesFund Size Vintage$650M USDTypical checkEarly-stage through growth, two dedicated fundsBest forFounders wanting AI-specific technical diligence and research accessLimitationAI-only mandate, no room for adjacent or pivoting theses Radical Ventures runs a two-tier AI-only structure that [founders should map before pitching](https://qubit.capital/startup-services/investor-mapping). The firm closed its fourth dedicated early-stage AI fund at [$650M USD, its largest early-stage vehicle to date](https://betakit.com/radical-ventures-holds-650-million-usd-final-close-for-new-early-stage-ai-fund/). It also runs a separate $800M growth fund that invests exclusively in later-stage AI companies. CPP Investments put $75M into the new fund and has committed $280M across Radical funds since 2019, a repeat-LP signal worth noting. Co-founder Jordan Jacobs built the Vector Institute with Geoffrey Hinton and sold his own AI startup, Layer 6, to TD in 2018. That is an operator credential, not a thesis written after ChatGPT launched. Hinton and Fei-Fei Li are LPs in Radical’s funds, alongside TD and Wittington Investments. What we seePartners here tend to push hard on compute cost assumptions before term sheets, not after. Expect questions on model differentiation that go past the pitch deck framing. Founders without a clear answer on defensibility against foundation-model incumbents get a slower process. ![NFX logo](data:image/png;base64,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)8. NFXFund Size Vintage$325M Fund IV closed October 2025Typical check$1M-$2M seed, via FASTBest forAI seed founders who want speed and certaintyLimitation9-day clock covers FAST track only NFX closed its [$325 million Fund IV](https://www.nfx.com/post/nfx-fund-iv) in October 2025, bringing assets under management to about $1.5 billion. The firm has backed nearly 200 startups since inception, including Lyft, DoorDash and Patreon, according to [coverage of the raise](https://www.calcalistech.com/ctechnews/article/bk2miixaxl). Fund IV explicitly names AI as a lead theme, alongside bioplatforms, space and security, per [reporting on the close](https://startupstorymedia.com/us-israeli-vc-firm-nfx-raises-325-million-to-fuel-ai-and-early-stage-startups/). The close also positions NFX among the largest early-stage funds globally, with roughly 50 new companies planned from the vehicle. NFX also runs a [nine-day FAST program](https://news.crunchbase.com/venture/nfx-speeds-up-seed-funding-with-new-nine-day-process/) for seed deals, promising an initial response within three business days. Checks under FAST run $1M-$2M for roughly 15% ownership, though that clock applies to FAST applications only, not partner-sourced deals. What we seeFAST’s software-driven intake means founders assemble their pitch mostly through a structured form, not a warm relationship. Diligence runs against standardized SAFE terms, so terms are less negotiable than at partner-sourced firms. Founders should treat the 9-day clock as a floor for the FAST track, not a guarantee for board-championed deals. ![Basis Set Ventures logo](data:image/png;base64,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)9. Basis Set VenturesFund Size Vintage$250M Fund IV announced 2026-01-26Typical checkPre-seed to Series A, low six figures up to $9M+Best forAI product, infrastructure, and robotics founders early, pre-narrativeLimitationFast-decision claims trace to third-party data, not Basis Set itself Basis Set Ventures closed a [$250M Fund IV](https://www.basisset.com/insights/announcing-our-250m-fund-iv) in January 2026, up from a $185M Fund III in 2024. Fund II closed at $165M in 2021, and the firm was founded in 2017. Founder [Dr. Lan Xuezhao](https://www.basisset.com/team-members/dr-lan-xuezhao) holds a Ph.D. in quantitative cognitive science from the University of Michigan and built Dropbox’s data science team. The firm backs [AI products, infrastructure software, and robotics](https://qubit.capital/blog/ai-mega-rounds-funding-trends) before the market narrative forms, according to [coverage of the Fund IV close](https://vcwire.tech/2026/01/28/basis-set-ventures-closes-fund-iv-at-250m/). Deal pace has stayed steady: 11 investments in 2025 and 8 more through May 2026, per [Tracxn’s portfolio data](https://tracxn.com/d/venture-capital/basis-set-ventures/__KMNSnpoXPyCefJr39GO6SULxUgUd0gC6sz1xQv5KW6I). One caveat: reports of 24-hour term sheets trace to a third-party fund database, not a Basis Set statement or founder account. What we seeBasis Set moves fast once a partner commits, but founders should verify decision timelines directly rather than lean on published averages. The firm’s studio support model means diligence conversations often cover how a founder plans to use post-investment introductions. 10. Conviction PartnersFund Size Vintage$230M Fund IITypical check$1M-$10M, up to Series ABest forSeed-stage AI-native foundersLimitationNo checks past Series A or above $10M Conviction Partners is a seed-stage investor capped at Series A, per [Ex-Greylock GP Sarah Guo Announces Conviction Partners, Her New AI Fund](https://www.forbes.com/sites/kenrickcai/2022/10/04/ex-greylock-gp-sarah-guo-announces-conviction-partners-her-new-ai-fund/). It raised a $230M Fund II in January 2025, per [Guo’s Conviction Partners adds Mike Vernal as GP, raises $230M fund](https://techcrunch.com/2025/01/31/guos-conviction-partners-adds-mike-vernal-as-gp-raises-230m-fund/). That fund is more than double the firm’s $101M debut from 2022. Mike Vernal joined as the second general partner that month, after nearly seven years at Sequoia Capital and eight at Facebook. Conviction has backed 27 startups, per [Sarah Guo Bet Everything On AI Pre-ChatGPT. Now She’s One Of The World’s Top Investors](https://www.forbes.com/sites/rashishrivastava/2026/05/27/sarah-guo-bet-everything-on-ai-pre-chatgpt-now-shes-one-of-the-worlds-top-investors/). Early bets include Harvey, OpenEvidence and Sierra, each worth billions by 2026. What we seeConviction’s six-of-27 board-seat ratio points to a hands-off style outside its core bets. Founders should confirm the Series A ceiling still holds before pitching. The larger Fund II could mean more flexibility than the original mandate suggests. 11. OpenAI Startup FundFund Size VintageSeed: 3 investmentsTypical checkDeal-by-deal, no fixed range; Series B rounds average ~$161MBest forLater-stage AI startups already inside OpenAI’s orbitLimitationNo new deals recorded in 2026; easy to confuse with the alumni fund OpenAI Startup Fund fits founders who want a backer tied directly into OpenAI’s product roadmap, per [Tracxn’s investor profile of the fund](https://tracxn.com/d/venture-capital/openai-startup-fund/__Rp6ugduZ613B4Hvba-352ySiK4qzAdIHtQFddJJlJW0). It has backed 14 companies total, with just 3 new investments in the year to November 2025, and none yet in 2026. Checks concentrate in Series B, averaging around $161M a round, far past typical seed and Series A range. Harvey rode an $80M Series B from the fund to an $8B valuation two years later, per [TechCrunch’s rundown of its portfolio](https://techcrunch.com/2025/03/01/openais-startup-empire-the-companies-backed-by-its-venture-fund/). A separate, OpenAI-alumni-run fund has begun investing under a similarly branded name, per [TechCrunch’s report on the alumni vehicle](https://techcrunch.com/2026/04/06/openai-alums-have-been-quietly-investing-from-a-new-potentially-100m-fund/). Partner Ian Hathaway brings an investing and economics background, not ML engineering, per [Global Venturing’s 2026 Powerlist profile](https://globalventuring.com/people/powerlist-2026-ian-hathaway-openai-startup-fund/). What we seeTerm sheets here tend to move through OpenAI’s own network rather than a standard partner meeting cadence. Expect diligence to probe model differentiation and compute cost assumptions harder than a typical seed check would. Confirm which fund, official or alumni-run, is actually signing before you get too far into the process. ![SignalFire logo](data:image/png;base64,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)12. SignalFireFund Size VintageInvested over 2.5 yearsTypical check$100K-$1M pre-seed, $1M-$5M seedBest forApplied AI founders, pre-seed to Series BLimitationWon’t fund foundation-model plays SignalFire is an AI-native venture firm built for founders raising pre-seed through Series B rounds in applied AI. The firm plans about 100 pre-seed checks of $100K to $1M, plus 60 seed checks of $1M to $5M, per [Crunchbase News](https://news.crunchbase.com/venture/signalfire-new-fund-seed-early-stage-applied-ai/). That capital deploys over roughly 2.5 years across pre-seed through Series B stages, per [TechCrunch](https://techcrunch.com/2025/04/07/signalfire-raises-over-1b-as-lps-embrace-data-driven-investing/). There is a catch. SignalFire deliberately avoids the foundation model layer, staying focused on applied AI with defensible data moats instead. Foundation-model founders should look elsewhere. Partner [Wayne Hu](https://www.signalfire.com/team/wayne-hu) has led seed-stage investing at the firm since 2015. His focus spans consumer bets and sectors ripe for end-to-end reinvention. In May 2025, SignalFire led a $50M Series B for Siro, an AI coaching platform for sales reps, per [TechCrunch](https://techcrunch.com/2025/05/21/siro-lands-50m-from-signalfire-to-expand-its-ai-powered-coaching-solution-for-on-ground-sales-reps/). What we seeSignalFire’s diligence tends to probe data moats early, since applied AI is the whole thesis. Expect questions about proprietary data access, not just model performance. Founders building at the foundation layer get filtered out fast in first calls. ## Why the Obvious Answer Is Wrong Most “top AI investor” lists rank firms by AUM or brand recognition, assuming bigger and better known means faster and friendlier. Our advisory work across AI-startup fundraises shows fund size does not predict velocity or diligence style. Some of the largest-AUM names on this list move slower and grill compute costs and model moat harder than smaller AI-dedicated funds. *From Qubit’s advisory work:* ‘Across 28 AI-startup fundraise advisory engagements from 2023 to 2026, founders who approached firms with a standing AI-only fund closed term sheets faster than those who led with generalist multi-sector funds writing AI checks out of a broader vehicle. AI-dedicated funds probed compute-cost assumptions and model-moat claims harder at diligence.’ Sequence outreach by fund structure, not fund size: AI-dedicated vehicles move faster but front-load tougher moat questioning. ### How to Spot a Real AI Investor Versus a Trend-Chaser Most firms added an “AI thesis” line to their website sometime in 2023. That page tells you nothing about whether they actually write checks in the space. Look at deal count instead. A firm with three or four AI deals a year, across multiple sub-sectors, is active. A firm with one portfolio logo and a lot of conference panels is not. Ask about their diligence questions in the first call. Investors who probe your compute costs or your model moat have done this before. Investors who only ask about total addressable market probably haven’t. ### Know Your AI Sub-Sector Before You Build the List Compute and infrastructure, foundation models, applied or vertical AI, and AI tooling are not the same fundraise. Each draws a different investor pool with different check sizes and different patience for burn. Foundation model investors expect long timelines and higher capital needs upfront. Applied and vertical AI investors move closer to how they’d evaluate any SaaS company, with revenue and retention mattering sooner. Figure out which bucket your startup sits in first. It narrows the list faster than any logo wall does. ## Conclusion This shortlist matched AI-focused investors to sub-sector and stage, named typical check size and pace. Separated firms actually deploying capital from those running an AI thesis on paper. That distinction is the difference between a warm intro and a wasted call. Sequence outreach by runway. Founders under nine months should start with the firms ranked earlier here, since they move fastest. Founders with more room to run can build a longer process around the later-ranked, larger-check firms. Ready to move? Qubit Capital’s [fundraising advisory](https://qubit.capital/startup-services/fundraising-assistance) helps AI startup founders build a runway-matched approach order and get warm positioning into the investors on this list.