· investment-strategies · 2 min read
NYC AI Startups Raised $23B from 2023–Q3 2025: The Data Behind the City's Applied AI Edge
NYC-based AI startups raised $23B across 2023 through Q3 2025 — but NYC's AI share is 7.5%, not 60% like SF. Here's why NYC wins applied AI despite the gap.
NYC AI has a different shape from Bay Area AI. Bay Area = frontier foundation models. NYC = applied AI for enterprise buyers. The numbers back this up.
The data
- $23B — NYC AI startup total funding, 2023 through Q3 2025 (NY State Comptroller).
- 35% — AI’s share of NYC capital raised in 2023 (Tech:NYC).
- 7.5% — NYC’s share of U.S. AI funding in 2024 (Comptroller).
- 60% — Bay Area’s share of global AI funding (2025 Bay Area Times).
- ~65% — AI’s share of total U.S. VC deal value in 2025 (Pitchbook/NVCA Q4 2025).
Why NYC “underperforms” on AI share
NYC looks small on AI percentage because:
- Foundation model rounds (OpenAI $122B, Anthropic $30B) are in the Bay Area and distort totals.
- Nvidia-ecosystem infrastructure deals cluster around Silicon Valley.
- Datacenter-scale capex (Meta, Microsoft) is a Bay Area + Texas + Washington story.
None of these “headline” categories reflect where NYC AI actually plays.
Where NYC AI wins
- Applied AI for finance: Hedge funds, asset managers, insurers — NYC is where the buyers are.
- Legal tech AI: Harvey, Dazzle, Casetext (pre-acquisition) — NY courts, NY law firms.
- Media + ad tech AI: Runway, Moonvalley, ElevenLabs. Sales to brands and publishers.
- Healthcare AI: Avo ($10M Series A), Patlytics, Faireez — NY hospital system density.
- Defense / dual-use AI: Palantir’s NYC hiring, DoD-adjacent NYC startups.
Notable 2025–2026 NYC AI deals
- Granola — $125M enterprise AI.
- Mirage — $75M AI video.
- Sycamore — $65M seed agent orchestration.
- Qodo — $70M code verification.
- Entire — $60M devtool seed.
- Sandbar — $23M Series A.
- Sequen — $16M personalization engine.
- Promptlayer — seed (Feb 2025).
- ElevenLabs — Series C (Jan 2025, NYC operations).
Why applied AI beats foundation models for most investors
- Faster revenue: Applied AI sells to known buyers; foundation models need massive capex.
- Better gross margin: Applied products don’t pay hyperscaler rent at scale.
- Defensibility through integration: Customer workflows and data are the moat.
- Exit optionality: Strategic buyers (banks, insurers, media) acquire applied AI more readily than foundation model companies.
Practical takeaway
- Founders: NYC is the best city in the U.S. for applied AI targeting finance, media, health, or legal.
- Investors: NYC’s AI share of dollars is lower but the portfolio-level IRR on applied AI often beats foundation model bets.
- LPs: Expect NYC AI exposure via applied verticals rather than GPU infrastructure.
Sources
- NY State Comptroller: https://www.osc.ny.gov/files/reports/osdc/pdf/report-13-2026.pdf
- Tech:NYC 2025 snapshot: https://www.technyc.org/nyc-tech-snapshot-2025
- Pitchbook-NVCA Q4 2025: https://nvca.org/wp-content/uploads/2026/01/q4-2025-pitchbook-nvca-venture-monitor.pdf
- vc-explainers
- nyc-startups
- ai
- +1 more