Accelerators: YC, Techstars and sector programmes, and when to join one

Core idea#

An accelerator is a short, intensive programme (usually three to six months) that invests a fixed amount in a cohort of startups in exchange for equity, and in return gives structure, mentoring, a peer group, and an investor showcase at the end. The best ones change a company's trajectory. For deeptech, they're useful, but only at the right moment.

What you get:

• Capital on standard terms, quickly, without negotiating.
• A forcing function: weekly goals and deadlines that push you to sell, ship and fundraise.
• A network: alumni, mentors and investors who take your calls.
• Signal: acceptance by a selective programme makes the next round easier. Sometimes facilities: labs and workshops at hard-tech programmes.

What it costs:

• Equity: usually 5% to 10%.
• Time and focus: three to six months, often relocating.
• Pace that may not fit deeptech: many programmes are built around software timelines, with a demo day where investors look for revenue growth.

The question isn't "is this a good accelerator?" It's "is this the right one, at this moment, for this company?"

Who it's for: deeptech founders who've incorporated (or are about to), have a team and a prototype, and are deciding whether to join an accelerator before raising a pre-seed or seed round.

Current terms#

Checked November 2026. Terms change; read each programme's own terms page before you apply.

Y Combinator#

• Deal: $500,000. $125,000 converts into a fixed 7%, and the other $375,000 goes in on an uncapped SAFE with "most favoured nation" terms, so it converts on the best terms you later give other investors (YC).
• Format: four batches a year (winter, spring, summer, fall) (F4 Fund), each about three months in San Francisco.
• Selectivity: acceptance is well under 1% of applicants on competitive batches (Causo).
• Deeptech fit: YC funds hardware, defence, energy and biotech, and its Requests for Startups now cover hardware, defence, agriculture and space (YesPress). Demo day still rewards visible progress; a deeptech company needs a strong technical milestone or early customer to stand out.

Techstars#

• Deal: $220,000: $20,000 through a post-money convertible equity agreement for 5% common stock, plus $200,000 on an uncapped MFN SAFE; no fee to join (Techstars).
• Format: three-month, mentor-driven programmes in many cities and themes (Techstars New York), including some focused on AI, robotics, health and deep tech (Slidebean).
• Deeptech fit: best where a city or sector programme matches your industry and its corporate partners are potential customers.

HAX (SOSV) — hard tech#

• Deal: a first investment of $250,000 that converts to a pre-negotiated equity stake later (HAX), with up to $550,000 advertised (Causo).
• Format: a six-month residency in Newark, New Jersey (Startup Grants India), in a facility with chemical, mechanical and electrical engineering labs, industrial 3D printers, metal fabrication and CNC machining (SOSV). HAX also has offices in Pune, Tokyo and Shenzhen (HAX).
• Deeptech fit: built for it: hardware, climate, industrial and health hardware teams with a prototype.

SOSV NY and SF (formerly IndieBio)#

Lab-based programme for science startups, now covering deep tech broadly; see playbook 1.10.

Sector and regional programmes#

• India: institute incubator acceleration programmes (playbook 1.11), corporate and sector accelerators (for example, in energy, mobility, agriculture and defence), and Entrepreneurs First Bangalore for pre-team founders (1.10).
• UK: sector accelerators linked to universities and catapults, Entrepreneurs First
• London, and deeptech venture builders.
• Global: Creative Destruction Lab (no equity, mentoring only) is a good complement to any of the above (1.10).

Comparison#

When to join, and when not to#

Join if most of these are true#

• You have a team (or are close) and a working prototype or strong technical result. You need help with commercial skills: sales, fundraising, hiring, more than technical help.
• The programme's network includes your customers or investors who back your sector.
• You can make visible progress in three to six months: a pilot, a technical milestone, first revenue.
• You'd otherwise struggle to raise a first round on good terms.

For hardware: the programme has the facilities you need (HAX-style).

Don't join (yet) if#

• The technology still needs 12 or more months of R&D before anything demo-able. A fellowship fits better (playbook 1.10).
• You can already raise a pre-seed round on better terms from investors who know your sector.
• The programme's investors don't fund your kind of company.
• Relocating would break your access to the lab, the university or your customers.
• Your IP isn't licensed or optioned yet (playbooks 2.03, 2.07). Most programmes need clean IP before they invest.

Do the dilution maths#

An illustrative example, assuming your next round is a priced seed at $20 million postmoney, and ignoring how the instruments interact in detail:

The accelerator money is more expensive per dollar than a good pre-seed round. You're paying the difference for the programme, the network and the signal. That's often worth it, but make the comparison explicitly. If your seed round is at a lower valuation, the MFN SAFEs take a larger share.

Worked example (illustrative)#

A three-person team building a robotic inspection system for wind turbines has a working prototype and two utilities interested, but no investors yet. Options:

1. HAX: facilities for the next prototype, a hardware-savvy investor, six months in Newark.
2. A Techstars energy programme: utility partners in the programme, three months, less cash.
3. A pre-seed round: one interested angel at a $6 million cap.

They choose the energy programme because its corporate partners are potential customers; it gets them a paid pilot, and they raise a seed round on the back of it. A team needing to build its next hardware iteration might rightly have chosen HAX instead.

Applying and getting the most out of it#

Before applying#

• Talk to two or three alumni from your sector. Ask what the programme did for them specifically, and what it didn't.
• Clean up the basics: incorporation, founder vesting (3.06), IP licence or option (2.07), a simple cap table.
• Prepare a short, clear application: one sentence on what you do; why now (1.09); evidence of progress; why this team.
• Record a short video if asked; keep it plain and direct.

In the interview#

Deeptech teams often lose interviews by spending the time on the science. Lead with the customer problem and progress, then the technology in one or two sentences, then the plan for the programme period. Be ready for: "What will you have achieved by demo day?" and "Why does this need venture money?"

During the programme#

1. Set two or three goals for the programme period on day one: a pilot signed, a technical milestone, a round committed.
2. Use the partners and mentors deliberately: book time early; bring specific problems.
3. Talk to customers every week. The programme pace is an opportunity, not just a pressure.
4. Start investor conversations before demo day. Demo day amplifies momentum you already have.
5. Don't let the programme pull you into a software playbook if it doesn't fit. Adapt the advice to your timelines.

After#

Alumni networks are often the longest-lasting value. Stay active: help later cohorts, ask for introductions, use the perks.

Scripts#

Asking an alumnus for an honest view#

"We're a [field] hardware team considering [programme]. You went through it in [year]. What did it change for you, specifically? And what would you warn a deeptech team about?"

The one-sentence description in an application#

"We make [product] that lets [customer] [outcome] at [X] the cost of [current method], based on [technology] developed at [lab]."

Answering "What will you achieve by demo day?"#

"Three things: a signed paid pilot with [customer], [technical milestone] demonstrated at [scale], and a seed round of [amount] committed. We're already in talks with [customer] and the milestone is [N] weeks of work."

Declining an offer politely#

"Thank you; we were genuinely impressed by the programme. We've decided to raise a pre-seed round with [sector] investors instead, because [reason]. We'd love to stay in touch, and may well apply to [later-stage programme]."

Common mistakes#

• Joining too early. With nothing to show by demo day, the signal works against you. Joining for the brand only. Fit with your sector, customers and investors matters more.
• Not modelling dilution, especially the MFN SAFE if your next round is small. Spending the interview on the science.
• Relocating and losing lab access you still need.
• Assuming demo day will raise the round. Investor conversations should start weeks before.
• Applying with messy IP or cap table. It can block the investment.

Related playbooks#

• 1.09 Founder-market fit
• 1.10 Fellowships for pre-company founders
• 1.11 Indian incubators
• 2.07 Licence anatomy
• 3.05 Splitting founder equity
• 3.06 Founder vesting

Sources#

• Y Combinator, The YC Deal; F4 Fund on YC batches; Causo H1 2026 accelerator report; YesPress on YC.
• Techstars, Investment terms update; Techstars New York; Slidebean on Techstars Boston.
• HAX, FAQ; SOSV, HAX Newark HQ; Causo on HAX; Startup Grants India on HAX.

Worked examples and dilution table are illustrative. Last checked November 2026.