What investors actually fund
What a16z funds at the early stage
Andreessen Horowitz’s seed and Series A bets, and its speedrun accelerator, turned into trends: which industries are breaking out, which spaces are crowded, and how each team’s companies have played out. Compare it with what YC funds.
Updated 2026-10-26 · refreshed monthly Download the numbers (CSV)Methodology
- Early-stage portfolio companies
- 672
- Speedrun companies (7 cohorts)
- 321
- Seed entry, 2023–2026
- 55%
- Fastest-rising team
- American Dynamism
For founders and investors
Breakout industries in speedrun
Industries whose share of speedrun companies grew most in the last two cohorts (SR006–SR007) compared with the two before (SR004–SR005).
Rising
Cooling
Barely existed two cohorts ago
Companies in the last two cohorts vs the two before.
For investors
Where a16z is putting new early-stage money
Each a16z team's share of the companies it first backed at seed or venture stage in 2023–2026, compared with 2019–2022. a16z adds new investments to its public list with a delay, so recent years are incomplete.
Rising
Cooling
For founders and investors
Trend any sector
Pick a speedrun industry or an a16z team and see how many companies it had in each cohort or year.
158 companies in AI Agents. SR006–SR007: 88 (63.8% of companies), vs 48 (47.5%) in SR004–SR005. Rising.
For founders
Crowded or open?
Speedrun companies per industry in the last 3 cohorts, compared with all earlier cohorts. 5 industries are crowded.
| Sector | Signal | Last 3 cohorts | Earlier cohorts | Share of cohort, SR001 → | Newest companies |
|---|---|---|---|---|---|
| AI Agents | Crowded | 125 ▲ | 33 | ||
| AI Infra | Crowded | 35 ▲ | 9 | ||
| Consumer | Crowded | 24 ▼ | 25 | ||
| Fintech | Crowded | 24 ▲ | 6 | ||
| Dev Tools & DevOps | Crowded | 20 ▲ | 7 | ||
| AI Models / Research | Active | 19 ▲ | 3 | ||
| Enterprise SaaS | Active | 18 ▲ | 3 | ||
| Media / Entertainment / Creator Economy | Active | 17 ▼ | 29 | ||
| Future of Work | Active | 16 ▲ | 3 | ||
| AI Generative Media | Active | 15 ▼ | 19 | ||
| Healthcare | Active | 14 ▲ | 6 | ||
| Robotics | Active | 13 ▲ | 4 |
For investors
How often a16z comes in at seed
Share of each year's new early-stage portfolio companies that a16z first backed at seed (the rest at venture stage).
For investors
How each team played out
Share of companies a16z first backed in 2021 or earlier that it lists as exited by IPO or acquisition.
- Acquired
- Public
- Active
- Enterprise46%
- Consumer45%
- Infra26%
- Bio + Health26%
- Games23%
- Fintech20%
- Cultural Leadership Fund14%
- Frontier3%
Companies a16z first backed in 2021 or earlier. a16z doesn't mark companies as shut down, so "active" means "not listed as exited". "Public" includes direct listings and SPACs.
For investors and founders
Where speedrun founders are based
1.3% of SR007 companies are based outside the US, down from 16% in SR001.
Share of each speedrun cohort based outside the US
Top cities
- San Francisco54.6%
- New York19.4%
- Los Angeles5.6%
- Palo Alto3.1%
- Mountain View1.5%
- London1%
For investors
Compare two a16z teams
Share of each year's new early-stage companies, 2012 onwards.
- Sector
- American Dynamism
- Enterprise
- Companies, all time
- 46
- 194
- First backed 2023–2026
- 20 (14.6%)
- 31 (22.6%)
- First backed 2019–2022
- 11 (3.6%)
- 60 (19.9%)
- Entered at seed
- 54%
- 41%
- Exited (backed by 2021)
- Too few
- 46% · 7 IPOs
- Newest
- Uplift, Belisar, Mariana Minerals
- Tessera Labs, Lio, Stuut
For founders
What a16z says it wants vs what it funds
a16z's Big Ideas 2026 lists what its partners want to fund. Next to each idea: the matching team's share of new early-stage companies.
American Dynamism
American Dynamism: 3.6% → 14.6% of new early-stage companies (2019–2022 → 2023–2026)
Bio + Health
Bio + Health: 15.2% → 21.2% of new early-stage companies (2019–2022 → 2023–2026)
Consumer Tech
Consumer: 19.2% → 10.9% of new early-stage companies (2019–2022 → 2023–2026)
Crypto & Web3
No matching team on the public list.
Enterprise + Fintech
Enterprise: 19.9% → 22.6% of new early-stage companies (2019–2022 → 2023–2026); Fintech: 13.9% → 5.8% of new early-stage companies (2019–2022 → 2023–2026)
Every speedrun cohort
a16z teams and speedrun industries
Methodology
Every number here is a count of companies a16z lists publicly. Nothing is estimated, and we never show dollar amounts: a16z doesn’t publish what it invested.
- Sources and freshness
- Two public a16z lists: its portfolio page and its speedrun accelerator directory. We store facts only (names, dates, stages, teams, industries, exits and location), never a16z’s descriptions, and every company links to its a16z page. Data as of 2026-10-26, refreshed monthly. Each refresh is reviewed by a person before it goes live.
- Which companies count
- The portfolio page is a selection of a16z’s investments, not all of them. We keep the companies a16z says it first backed at seed or venture stage, and leave out growth-stage entries. Years are the year of a16z’s first investment, from 2012. a16z adds new investments to the list with a delay, so the most recent years look smaller than they are. Speedrun counts every company in each cohort a16z lists.
- Teams and industries
- Teams are a16z’s own investing groups (for example Enterprise, Bio + Health, American Dynamism). A company can belong to more than one. Industries are the labels speedrun companies carry in a16z’s directory, and a company can have several.
- Breakout themes
- Speedrun industries compare each industry’s share of companies in the last 2 cohorts with the 2 before. An industry needs 8+ companies in the recent cohorts to rank. “Barely existed before” means 6+ recent companies and at least three times the earlier count. Teams compare their share of new early-stage companies in 2023–2026 with 2019–2022; four-year windows, because recent years are thin.
- Trend any sector
- Counts per cohort (industries) or per year (teams). “Rising” means the recent share is at least 25% higher than before; “cooling” at least 20% lower. With fewer than 8 companies across both windows we don’t call a trend.
- Crowded or open
- From speedrun companies in the last 3 cohorts. An industry is Crowded with 20+ companies, Active with 8–19, and Open below 8. The arrow compares the per-company rate with all earlier cohorts (up or down by more than 15%). “Open” describes speedrun only, not the whole market.
- How each team played out
- For companies a16z first backed in 2021 or earlier (at least five years ago): the share a16z lists as exited by IPO (including direct listings and SPACs) or by acquisition. A team needs 20+ such companies. a16z doesn’t mark companies as shut down, so “active” means “not listed as exited”. This shows how often companies exited, not how much anyone made.
- Where speedrun founders are based
- The city and country each speedrun company lists. The portfolio page has no locations.
- Download
- Every number on these pages is in the CSV download. You’re welcome to cite it; please link to this page.
Frequently asked questions
What does a16z invest in at the early stage?
Among the 672 companies a16z lists as first backed at seed or venture stage, the largest teams are Enterprise (194), Consumer (144), Bio + Health (101). The fastest-growing is American Dynamism: 3.6% of new early-stage companies in 2019–2022, 14.6% in 2023–2026.
What is a16z speedrun?
Speedrun is a16z's accelerator for very early startups, run in cohorts. Its public directory lists 321 companies across 7 cohorts (SR001–SR007), growing from 25 to 79 companies per cohort.
What kind of startups does a16z speedrun fund?
AI Agents leads: 63.8% of companies in the last two cohorts. The most common industries across recent cohorts are AI Agents, AI Infra, Consumer, Fintech. SR007's median team size is 3, down from 12 in SR001.
How often does a16z invest at seed?
55% of the early-stage companies a16z first backed in 2023–2026 came in at seed; the rest at venture stage (Series A and later). a16z adds new investments to its public list with a delay, so recent years are incomplete.
Looking for what to build next?
See the open problems on Questd that investors want solved, and how YC’s bets compare.