What are the stages of startup funding, from pre-seed to Series A?
Rounds are named by what the company has proved, not by a fixed amount. Pre-seed backs a team and an early product, often from angels or a program: YC invests $500,000, Antler India INR 4 crore for 11%. Seed backs early traction while the company finds a repeatable way to grow: Surge invests $500,000 to $5 million, and Blume writes $1 to 1.5 million for 15 to 20%, sized to last 15 to 18 months. Series A backs a business with product-market fit that needs capital to scale, usually led by an institutional VC taking a board seat. Each priced round sells roughly 20%, the norm Carta's data bears out.
This is the actual math you are trading against, straight from the source: $125K for a fixed 7%, plus a $375K uncapped MFN SAFE that converts at your next round's lowest cap. Read it before you romanticize the check. On a $15M cap that SAFE alone is another 2.5%, so YC ends up owning roughly 10% of your company for $500K. Indian founders should also note the fine print: YC only invests into US, Canada, Cayman, or Singapore entities, so an Indian company must flip its parent offshore to take the deal.
Written by a partner at Blume Ventures, this is the India-specific version of exactly what we're telling founders: valuation is not a spreadsheet output, it is (in Fred Wilson's line the piece builds on) the exhaust fumes of a negotiation about two things, the amount raised and the dilution. It shows the real Indian seed math (a roughly 1 to 1.5M dollar cheque for a 15 to 20 percent stake, back-solving to a 5 to 7.5M dollar post-money) and why the VC sizes the round to get you to Series A, not to please your ego.
This is the cleanest statement of the per-round rule of thumb backed by Carta's data across 1,200+ rounds: roughly 20 percent at seed, 20 percent at Series A, 15 percent at Series B, then 10 to 15 percent. Lemkin's blunt closing point (that dilution adds up and there are real benefits to being efficient instead of chasing every round) is exactly our stance. Do the arithmetic on those numbers and two founders splitting the company are already near the single digits by Series B if every round runs hot, which is why you model forward before you sign.
The canonical, no-nonsense overview of how seed fundraising works, written by a former YC president who has seen thousands of rounds. It's the single best starting point before you talk to a single investor.
This is the cleanest decision framework we found because it draws the line by cheque size, not vibes: under 1M and 1M to 2M go SAFE, 2M to 5M model both, above 5M go priced (70% of those deals are). It also names the trap our answer warns about, that every new SAFE dollar dilutes only you and stacking creates compounding asymmetric dilution you won't feel until Series A conversion.
Round size is the deciding variable: SAFEs dominate below 2M, priced rounds dominate above 5M, and 2M to 5M is where you actually model both side by side
A priced round buys the lead their liquidation preference, protective provisions, pro rata, information rights, and usually a board seat, which is the real reason a big-cheque lead wants it
Before signing anything, model the full conversion including option pool expansion, because SAFE stacking hides dilution until it converts
Antler's India page, for founders at day zero, including solo founders still looking for a co-founder. It sets out the AI Residency terms: a decision in three weeks and INR 4 crore (about $470K) for 11%, and notes that about half its India investments are made directly outside the program.
Peak XV's early-stage program, the most selective seed program run from India. Its page states it invests from $500,000 up to $5 million in seed capital, with operators embedded across hiring, product, GTM and expansion.
