The US Accelerator Landscape: Where to Apply and What to Stack

Last month a founder asked us whether they should apply to YC, Techstars, and three credit programs simultaneously. The answer was yes to YC, no to Techstars (wrong stage), and yes to all three credit programs. The credit programs don't take equity. You apply once and forget about them.

The US accelerator world has split cleanly: programs that write checks and take equity (YC, Techstars, 500 Global, AI Grant, AI2, Antler) and programs that give you compute credits and ecosystem access without dilution (NVIDIA Inception, Microsoft Founders Hub, Google for Startups, AWS Activate). Most founders should be in one of the first group and two or three of the second.

This is the working list of what's worth applying to, what each program actually gives you, and how to decide.

Y Combinator

Investment: $500K total — $375K on an MFN SAFE, plus $125K on a SAFE with a $1.8M post-money cap.

Equity: 7% on the $125K piece, plus whatever the $375K converts to at your next round.

Program length: 13-14 weeks.

Cohort size: ~125 companies per batch. Four batches a year as of 2025.

Acceptance rate: ~1%.

Best for: First-time founders, pre-product, B2B SaaS, dev tools, AI infrastructure.

YC is the default choice if you get in. The brand still opens doors that other accelerators don't. The network is unmatched. The acceptance rate is brutal, but the application is short enough that you should apply unless you're certain you won't get in.

We wrote a longer framework on whether YC makes sense for your company.

Techstars

Investment: $220K total — $20K up front, $200K SAFE.

Equity: 5%.

Program length: 13 weeks.

Locations: 11+ cities globally, with vertical-specific cohorts (fintech, healthcare, climate).

Acceptance rate: 1-2% per cohort.

Best for: Founders who need a local mentor network or are in a vertical with a dedicated Techstars program.

Techstars trades brand prestige (lower than YC) for vertical depth and a smaller equity ask. The mentor model is more hands-on than YC's. If you're in Boulder and building climate tech, Techstars might be better than YC. If you're in SF building a horizontal SaaS tool, probably not.

500 Global

Investment: $150K standard.

Equity: 6%.

Program length: 4 months.

Best for: International founders, emerging markets, founders building for the US market from outside it.

500 Global has more international reach than YC or Techstars. If you're a non-US founder who needs a US entity and introductions to US investors, this is the strongest option. The LP network skews global, which helps if you're raising a Series A outside Silicon Valley.

AI Grant

Investment: $250K standard.

Equity: Variable — usually a small SAFE on a high cap.

Program length: Lighter touch than YC. Mentor-driven, not cohort-bound.

Founded by: Daniel Gross and Nat Friedman.

Best for: AI-only startups with a strong technical thesis.

Highly selective. The Daniel Gross / Nat Friedman network is uniquely valuable if you're building AI infrastructure or working on frontier model problems. This is not a program for "we use GPT-4 in our product" startups. It's for people building the next layer of the stack.

AI2 Incubator

Investment: Variable, often $500K-$1M.

Equity: Variable.

Best for: AI research-backed startups, deeply technical founders, pre-product.

AI2 is tied to the Allen Institute for AI. If your startup is coming out of an academic lab or you're working on research-heavy AI problems, this is worth applying to. Less relevant if you're building an application layer product.

Antler

Investment: $100-200K pre-formation, then participation rights in follow-on rounds.

Best for: Founders who don't have a co-founder yet.

Antler runs a 10-week co-founder matching process, then invests in the teams that form. If you're solo and don't have a technical co-founder, this is the only program that solves that problem structurally. If you already have a team, skip it.

Credit programs (no equity, stack with everything)

These don't take equity. Apply to all of them.

NVIDIA Inception — Compute credits, technical access, and introductions to NVIDIA's enterprise sales team if you're building infrastructure.

Microsoft Founders Hub — $150K in Azure credits plus $2.5K in OpenAI API credits.

Google for Startups — Cloud credits and mentorship. Less structured than Microsoft's program, but the credits are real.

AWS Activate — Up to $100K in AWS credits depending on your stage and backing.

Plug and Play — Corporate connections in 60+ locations. Variable structure, mostly useful if you're selling to enterprise and need warm intros.

If you're building anything that uses compute — and especially if you're training models or running inference at scale — these credits are worth $50K-$150K in real cost savings. Apply to all of them. The applications take 20 minutes each.

Decision matrix

Should you apply to multiple?

Yes, but with discipline.

Apply to one or two cash-providing programs that are a genuine fit. Apply to two or three credit programs regardless. Do not apply to five accelerators at once. The applications take 2-4 hours each, and acceptance signals leak between programs. YC and Techstars partners talk.

If you're pre-formation and don't have a co-founder, Antler is uniquely valuable. If you're building AI infrastructure, AI Grant plus NVIDIA Inception is the strongest combination. If you're an international founder targeting the US market, 500 Global is worth prioritizing over Techstars.

For more on whether YC makes sense for your company, see our YC framework piece. Book a call if you want a second opinion on accelerator strategy.