An analysis of 1,129 deals, $26.3 billion in capital, and the market dynamics that headlines miss.
When xAI raised $20 billion in a single round and OpenAI hit a $500 billion valuation, the message seemed clear: AI funding is a game of superlatives. Go big or go home.
But buried beneath the mega-round headlines is a different story. One told by 1,129 Series A deals totaling $26.32 billion in 2025.
The median deal? Just $16 million.
That’s not a typo. In a year where foundation models absorbed $80 billion and 58% of AI funding flowed to $500M+ mega-rounds, the typical Series A AI company raised a disciplined round that would barely cover two months of compute costs for a frontier lab.
This report goes beyond the numbers we published in our newsletter. We analyzed all 1,129 deals across 264 cities, 15 sectors, and 20+ technology categories to understand what Series A AI funding actually looks like — who’s investing, where the money goes, and what the data says about where the market is heading.
On this page
- The Mega-Round Mirage
- Where the Money Actually Goes: Deal Size Distribution
- The Quiet Majority: 84% of Deals Happen Outside San Francisco
- The Second City Premium: Boston vs. London
- Sector Deep Dive: Where AI Capital Deploys
- Technology Trends: The AI Agents Supercycle
- The AI Stack: Applications vs. Infrastructure
- Quarterly Seasonality: When Money Moves
- The Investor Landscape: Who's Writing Checks
- Five Shifts for Investors and Founders
- Key Takeaways
- Methodology
The Mega-Round Mirage
Media coverage of AI funding has created a distorted picture of what raising Series A actually looks like. When every headline features nine-figure rounds, founders internalize a false benchmark.
The distortion by numbers:
- Foundation models captured $80B in 2025 (40% of all AI funding)
- 58% of total AI funding went to $500M+ mega-rounds
- Top 3 deals (xAI, OpenAI, Anthropic) absorbed roughly $50B alone
- Yet the median Series A was $16M — 1,250x smaller than xAI’s raise
The consequence? Founders either chase unrealistic valuations, feel inadequate raising “only” $15-25M, or over-concentrate on San Francisco assuming geography determines fundability.
Series A remains the last stage where traditional venture economics apply. Price discovery works. Fundamentals matter. And 1,129 companies proved it.
Where the Money Actually Goes: Deal Size Distribution
The $16M median masks an informative distribution. Two-thirds of all Series A AI deals fell into a single band — and it tells us exactly what “normal” looks like at this stage.
| Deal Size | Deals | Funding | Share of Capital |
|---|---|---|---|
| < $1M | 4 | $2.7M | 0.0% |
| $1M – $5M | 61 | $188M | 0.7% |
| $5M – $10M | 159 | $1.14B | 4.3% |
| $10M – $50M | 755 | $15.1B | 57.6% |
| $50M – $100M | 68 | $4.2B | 16.0% |
| > $100M | 36 | $5.6B | 21.4% |
The $10-50M band is where Series A lives. It accounts for 66.9% of all deals and 57.6% of capital deployed. The 36 deals above $100M attracted outsized attention but represented just 3.2% of deal count.
This creates two distinct economies within a single funding stage. The disciplined majority — companies raising $10-50M with realistic 18-24 month runways at $700K-900K monthly burn. And the outlier class — 36 companies averaging $156M per round, facing impossible math where they need $500M+ Series B rounds to avoid down rounds.
As Anders Ranum, Partner at Sapphire Ventures, declared: “2026 is a fundamentals-first year where capital rewards revenue growth, efficiency and real AI advantage, and punishes anything that is AI veneer on old ideas.”
The $16M median isn’t a ceiling. It’s a sanity check.
The Quiet Majority: 84% of Deals Happen Outside San Francisco
951 deals happened outside San Francisco. That’s 84.2% of Series A activity — the quiet majority that headlines ignore.
| Metro | Deals | Total Funding | Avg Deal Size |
|---|---|---|---|
| San Francisco | 178 | $4.15B | $23.3M |
| New York City | 132 | $2.88B | $21.8M |
| London | 58 | $1.23B | $21.2M |
| Boston | 38 | $1.15B | $30.3M |
| Tel Aviv | 26 | $672M | $25.8M |
| Los Angeles | 23 | $274M | $11.9M |
| Paris | 22 | $610M | $27.7M |
| Seattle | 18 | $357M | $19.8M |
| Singapore | 18 | $325M | $18.0M |
| Austin | 16 | $632M | $39.5M |
The paradox: 65%+ of AI-focused VC capital is managed by SF-based firms, yet 84% of deals happen elsewhere. This creates a structural inefficiency — SF VCs must compete intensely locally or travel extensively for non-local deals.
The Second City Premium: Boston vs. London
Boston: Conviction Capital
- 38 deals, $1.15B total
- $30.3M average — highest of major metros, 30% higher than SF
Boston’s biotech-AI convergence attracts drug discovery, computational biology, and clinical AI — categories requiring larger rounds due to regulatory timelines and data requirements. MIT/Harvard spinouts have established credibility. Flagship, Atlas, and Third Rock write larger checks for longer holds.
Notable deals: Layer Health ($21M, clinical decision support), 7AI ($130M, cybersecurity), OpenHands ($18.8M, AI coding agents).
London: Experimentation at Scale
- 58 deals, $1.23B total
- $21.2M average with 53% more deals than Boston
US companies often establish London offices for European expansion, captured in UK deal flow. SEIS/EIS tax advantages make early-stage funding more accessible. The DeepMind legacy plus Imperial/UCL/Cambridge create continuous startup formation.
Notable deals: Ultralytics ($30M, YOLO computer vision), nsave ($18M, fintech infrastructure), Fyxer AI ($10M, 20VC-backed).
Both work. Similar total funding ($1.15B vs. $1.23B) suggests neither approach is strictly superior. Founders should match their strategy to market.
Sector Deep Dive: Where AI Capital Deploys
Series A AI funding isn’t a monolith. It’s 15 distinct markets, each with different dynamics, deal sizes, and investor profiles.
Health & Biotech leads in capital deployed ($7.71B across 230 deals) with an average deal of $33.5M — the highest of any major sector. Drug discovery and precision medicine rounds run larger because regulatory timelines demand longer runways.
Enterprise Software leads in deal count (279 deals, $5.79B) with a more modest $20.7M average. This is the broadest category, spanning AI agents, document automation, and data management.
FinTech holds the middle ground (192 deals, $4.02B) at $20.9M average. Compliance automation and risk assessment drive deal flow.
The Funding Efficiency Gap
Five sectors are systematically underfunded relative to the market average of $24.3M:
| Sector | Avg Deal | vs. Market Avg | Severity |
|---|---|---|---|
| Education Technology | $14.1M | 58% of average | Moderate |
| Retail & E-Commerce | $14.1M | 58% of average | Moderate |
| Food & AgriTech | $14.9M | 61% of average | Moderate |
| Real Estate & Construction | $15.8M | 65% of average | Moderate |
| Marketing & Sales Tech | $17.0M | 70% of average | Moderate |
These underfunded sectors represent opportunity for investors willing to go against consensus. The gap between market size and funding intensity signals either a timing mismatch or structural bias in VC allocation.
Technology Trends: The AI Agents Supercycle
AI Agents commanded 257 deals and $6.07B — more deals than any other technology category by a wide margin. This reflects the industry’s pivot from “AI that generates” to “AI that acts.”
But the most capital-efficient tag tells a different story. Biotechnology attracted just 91 deals but commanded $4.53B — an average of $49.8M per deal, more than double the market average. Drug Discovery (64 deals, $4.05B) and Therapeutics (67 deals, $3.92B) follow the same pattern: fewer deals, larger checks, longer conviction horizons.
| Tag | Deals | Total Funding | Avg per Deal |
|---|---|---|---|
| Drug Discovery | 64 | $4.05B | $63.2M |
| Biotechnology | 91 | $4.53B | $49.8M |
| Therapeutics | 67 | $3.92B | $58.5M |
| Precision Medicine | 62 | $2.58B | $41.7M |
| Robotics | 60 | $2.44B | $40.7M |
| Industrial Automation | 111 | $3.04B | $27.4M |
| Payments | 64 | $1.71B | $26.7M |
| AI Agents | 257 | $6.07B | $23.6M |
| Threat Detection | 64 | $1.40B | $21.9M |
| Supply Chain | 100 | $2.01B | $20.1M |
The AI Stack: Applications vs. Infrastructure
AI Applications: 914 deals (79.2%), $20.8B total. AI Native: 580 deals, $13.6B (companies that couldn’t exist without AI). AI Augmented: 334 deals, $7.2B (existing products enhanced with AI).
AI Infrastructure: 167 deals (20.8%), $5.5B total. AI Adjacent: 104 deals, $2.8B (enabling AI but not running models). AI Platforms: 63 deals, $2.6B (building foundation models/chips).
The infrastructure layer captures disproportionate funding per deal. AI Platforms average $41.9M per deal versus AI Native’s $23.4M. Infrastructure rounds run larger because the capital requirements — compute, data pipelines, specialized hardware — demand it.
Vertical vs. Horizontal: Vertical AI (industry-specific) captured 880 deals (81.4%), while Horizontal AI (cross-industry) had 203 deals (18.6%). The overwhelming preference for vertical AI at Series A makes sense — investors at this stage want clear markets, identifiable customers, and measurable revenue.
Quarterly Seasonality: When Money Moves
| Quarter | Deals | Funding | Median Deal | QoQ Change |
|---|---|---|---|---|
| Q1 2025 | 286 | $6.96B | $15.2M | — |
| Q2 2025 | 271 | $5.82B | $15.0M | -16.4% |
| Q3 2025 | 273 | $7.15B | $16.0M | +23.0% |
| Q4 2025 | 253 | $6.38B | $17.0M | -10.8% |
Two patterns stand out. First, deal count was remarkably stable — between 253 and 286 deals per quarter, never varying more than 13%. Second, median deal size crept upward from $15.2M in Q1 to $17.0M in Q4 — a 12% increase suggesting gradual pricing inflation.
For founders, the implication is tactical: avoid raising in August, target September-November or January-March for optimal market conditions.
The Investor Landscape: Who’s Writing Checks
Y Combinator dominates with 67 deals ($1.41B in deal value), functioning as the de facto pipeline for Series A AI companies. Six of the top 13 co-investment pairs involve YC as one partner.
The tier below YC shows a five-way tie at 20-21 deals: Andreessen Horowitz (21 deals, $650M), General Catalyst (20 deals, $1.03B), Lightspeed (20 deals, $921M), Accel (20 deals, $635M), and Insight Partners (20 deals, $451M).
Five Shifts for Investors and Founders
1. The Median Is the Message
Stop using mega-rounds as benchmarks. A $16M Series A with 3x growth to Series B is a success. Companies raising $16M face realistic expectations: 18-24 month runway, clear metrics for progression. Companies raising $100M+ face impossible math.
2. Geographic Arbitrage Is Real
951 deals outside SF represent structural deal flow opportunity. An SF AI engineer costs $350-500K fully loaded. Boston: $250-350K. London: $180-280K. The same $16M buys 40-60% more runway outside the Valley.
3. Vertical AI Is the Series A Sweet Spot
81.4% of deals targeted specific industries. At the $16M level, vertical focus — identifiable customers, measurable revenue, clear competitive moats — is what investors fund.
4. The Underfunded Sectors Are Your Opportunity
Education Technology, Retail/E-Commerce, Food/AgriTech, and Real Estate all raise 35-42% below market average. These sectors have massive TAMs but lack dedicated AI capital. For investors, this is where contrarian bets live.
5. The Series B Bottleneck Is Coming
1,129 Series A companies will compete for Series B in 2026-2027. Series B capacity hasn’t scaled proportionally. The discipline advantage compounds — companies that raised $16M with realistic expectations have options when Series B tightens.
Key Takeaways
For Founders
- The typical Series A AI round is $16M, not $100M. Build your plan accordingly.
- 84% of deals close outside San Francisco. Location matters less than execution and customer proximity.
- Target September-November or January-March for optimal fundraising conditions. Avoid August.
- Vertical AI (81% of deals) is what Series A investors fund. Prove your market before going horizontal.
For Investors
- Y Combinator’s 67-deal pipeline makes it the essential Series A feeder. Build or maintain YC relationships.
- Five sectors (EdTech, Retail, AgriTech, Real Estate, MarTech) are systematically underfunded at 58-70% of market average.
- Boston ($30.3M avg) and Austin ($39.5M avg) produce the largest deals. London (58 deals) produces the most volume outside the US.
- The AI Applications layer (79% of deals) is where volume lives, but AI Infrastructure (21%) captures higher per-deal economics.
For the Market
- Series A is the last rational stage in AI — pricing discipline still holds at $16M median.
- 264 cities participated in AI Series A funding in 2025. The geographic distribution is far wider than narrative suggests.
- Biotech/drug discovery commands 2-3x the per-deal funding of software categories. Two parallel markets exist within Series A.
- Q4 2025’s rising median ($17M) signals gradual pricing inflation heading into 2026.
Methodology
Analysis Period: January 1 — December 31, 2025
Data Source: Bot Memo proprietary database. 1,129 Series A AI funding rounds with disclosed amounts.
Inclusion Criteria: Announced funding amount from disclosed rounds. Company self-identifies as AI/ML-focused or uses AI in core product. Series A designation confirmed by company or lead investor.
Exclusion Criteria: Undisclosed amounts not estimated. Debt financing, grants, and ICOs excluded. Secondary transactions excluded. Series A extensions counted with original round unless separately announced.
Key Metrics: Median calculated across all 1,129 disclosed deals. Averages calculated as total funding divided by deal count per geography. Metro assignment based on company HQ at time of announcement.
Limitations: Undisclosed rounds not included — may skew toward larger disclosed deals. Multi-HQ companies assigned to primary location. Currency converted to USD at announcement-date exchange rates.
This analysis is based on Bot Memo’s proprietary database of 1,129 Series A AI funding rounds in 2025. Bot Memo catalogs AI startups worldwide to surface investable market gaps for VCs and founders. Subscribe to our newsletter for weekly analysis.
About the author
Chintan Zalani
I'm the insight architect behind Bot Memo. I have spent the last decade building media assets on the internet. Bot Memo started as a simple project covering industry deep dives. Then I built a data pipeline for it. And now I love analyzing and covering all things AI startups and trends on top of our own data infrastructure.
