How to apply to a16z Speedrun in 2026

a16z Speedrun gives accepted founders up to $1M ($500K for 10% upfront via SAFE, plus $500K in the next round) and runs a 12-week San Francisco cohort twice a year. Acceptance is below 0.4%. This guide breaks down the 5 application questions, the deal terms, the SR007 timing, and how to position against the tech-meets-entertainment thesis instead of writing a generic YC apply.

Table of contents

• The 5 a16z Speedrun application questions
• What a16z Speedrun actually funds in 2026
• The a16z Speedrun deal terms, line by line
• How to position for the Speedrun thesis
• SR007 dates, deadlines, and what happens next
• a16z Speedrun vs Y Combinator: where each one wins
• What Speedrun acceptance actually requires
• Who should not apply to a16z Speedrun
• FAQ

Most founders learning how to apply to a16z Speedrun copy the Y Combinator playbook, which is the single biggest reason they get rejected. Speedrun is not a smaller YC. It is a thesis-driven program with a fixed deal, a specific entertainment-and-AI mandate, and a sub-0.4% acceptance bar (a16z Speedrun FAQ). If you write to "we are building a startup," you are competing against thousands of generic applicants. If you write to a16z's actual thesis (tech crossed with entertainment, AI, 3D, creative tools, gaming, streaming, XR), you are writing to two human partners who already want to fund what you are building. This guide walks the five application questions, the deal terms, the SR007 dates, and the positioning calls that survive partner review.

The 5 a16z Speedrun application questions

The a16z Speedrun application is a short text form with five questions, reviewed by at least two human partners with no automated screening (a16z Speedrun , FAQ). Here is what each question is testing and how to answer it.

1. What are you building? One paragraph that names the product, the surface area, and the wedge. Lead with the noun ("a TikTok-clip-to-podcast tool that…"), not the verb ("we are building…"). Partners scan the first sentence and decide whether to keep reading.
2. Who is it for? Name a specific user, not a market. "Twitch streamers with 1K–10K followers monetising on subscriptions" beats "creators." Specificity is the signal that you have done the customer work.
3. Why now? State the technical, behavioural, or market shift that makes this product possible in 2026 and not in 2024. Generative video unit cost, a new model context window, the closure of a competitor: any one is a real why-now. "AI is changing everything" is not.
4. Why you? Cite the specific experience that makes you the right team. Operator background, technical credentials, prior projects with traction. Two human partners want a one-sentence story of why you, in particular, will out-execute the other 0.4%.
5. What is working? Hard numbers. Daily active users, retention, revenue, partner conversations, waitlist signups. Anything qualitative ("users love it") is read as "we have no traction."

The single most common rejection cause is treating these five as a free-form mini-pitch. Each is a specific question with a specific answer shape. Treat them as such, and the Speedrun acceptance odds shift in your favour.

What a16z Speedrun actually funds in 2026

Speedrun funds tech crossed with entertainment, not games-only and not generalist B2B. Since launching in 2023, the program has deployed more than $300M to over 250 startups, with nearly 10,000 investors engaged with companies in the last batch alone. The March 2025 rebrand made the expanded thesis explicit: AI, 3D, creative tools, gaming, streaming, XR, and next-gen entertainment all sit inside the mandate (a16z , Meet the new a16z Speedrun).

If your startup has no plausible entertainment, AI, or creative-tools angle, apply elsewhere. The acceptance bar is brutal for off-thesis founders because the partners reviewing your app have a specific brief, and a brilliant fintech infrastructure pitch is not it. The Speedrun accelerator is built for founders who can sit inside one of those buckets without contortion.

Worth knowing: aggregator listicles still describe Speedrun as a games-only program. They are stale. The post-rebrand thesis is broader, which means the a16z Speedrun 2026 application pool is broader too. Generative AI startups, creator-tooling companies, and XR hardware founders all show up in the cohort.

The a16z Speedrun deal terms, line by line

Speedrun has a single standard deal for every accepted company, with no negotiation. The a16z Speedrun deal terms break down like this:

• $500K upfront for 10% common, via SAFE: the initial check, taken into a post-money SAFE at the time of acceptance.
• $500K follow-on in the next round: invested within 18 months on the same terms as your next lead investor (a16z Speedrun , The Deal).
• Pro rata rights in subsequent financings: a16z keeps the option to maintain ownership in your A and B.
• $7M+ in partner credits across 250+ tools: redeemable across the cohort, from cloud compute to design tools (a16z Speedrun , Apply).
• 600+ Speedrun alumni network: alumni community access for current and accepted founders.
• No board seat at this stage: a16z explicitly does not take a board seat at the Speedrun stage (a16z Speedrun , The Deal).
• 12-week cohort in San Francisco: in-person, twice per year (a16z Speedrun , FAQ).

The implied valuation is a $5M post-money cap on the upfront tranche. That is below the 2025 market median for pre-seed: Carta's State of Pre-Seed 2025 reports median post-money SAFE caps of approximately $10M for $250K–$1M rounds and $15M for $1M–$2.5M rounds (Carta , State of Pre-Seed 2025). You are accepting a below-market cap in exchange for the brand, the credits, and the demo day exposure.

Don't treat the Speedrun deal as a benchmark for your next round. The $500K-for-10% is a program-level price, not a market signal. When you talk to seed leads after demo day, anchor on the post-Speedrun round valuation, not the Speedrun cap.

How to position for the Speedrun thesis

The biggest application mistake is writing for a generalist investor when a16z Speedrun 2026 has an explicit, narrow thesis. The fix is to map your startup to one of the thesis buckets in the first sentence of "what are you building." If the partner reading your app cannot put you into a bucket (AI for entertainment, creative tools, 3D, gaming, streaming, XR, generative content), you lose them in 30 seconds.

Two examples on the "why now" question, where positioning lands hardest:

✅ Good: "Generative video crossed the latency-and-cost threshold to drop a per-clip cost from $4 to $0.06 in the last six months, which is why our Twitch-clip-to-TikTok tool now works." Why it works: names a specific technical shift, with a specific number, mapped to a specific product opportunity inside the thesis.

❌ Bad: "AI is changing everything, and creators are looking for new tools to take advantage of the AI revolution." Why it fails: generic, untimed, and unspecific to entertainment. The partner has read this opener 500 times in the same batch.

For "who is it for," anchor on a creator, player, viewer, or audience archetype. "Independent game studios shipping on Steam" maps to the gaming thesis bucket. "Series A SaaS companies optimising their funnel" does not, even if it is true and even if the product is good.

Don't apply with a generic B2B positioning and hope the partners infer the entertainment angle. They will not. Make the bucket explicit in the first 100 characters of your application.

SR007 dates, deadlines, and what happens next

The SR007 cohort runs July 27 to October 11, 2026 in San Francisco, with Demo Day on October 6, 2026. The application deadline was May 17, 2026 at 11:59 PM PT (a16z Speedrun , SR007 Announcement). Speedrun runs twice a year, so the next open application cycle after SR007 closes will be SR008, opening for review during the second half of 2026.

What happens after you submit:

• Two human partners review the application: no automated screening, no algorithmic gating (a16z Speedrun , FAQ).
• Shortlist interview: a short partner call within roughly two to six weeks of submission, depending on volume.
• Final partner review: a longer working session with at least one of the named Speedrun partners.
• Decision and offer: standard SAFE paperwork sent to accepted founders.
• Cohort kickoff: in-person in San Francisco on the published start date.

Don't apply with one week to go before the deadline. Late applications get reviewed last and compete against the freshest pool of applications partners have seen. Get yours in during the first two weeks of the open window.

a16z Speedrun vs Y Combinator: where each one wins

The Speedrun vs YC comparison is the highest-volume adjacent search to how to apply to a16z Speedrun, and it usually gets covered badly. Here is the side-by-side that matters.

Sources: a16z Speedrun , The Deal, a16z Speedrun , FAQ.

Apply to Speedrun if your startup sits cleanly inside one of the thesis buckets and you want the a16z brand alongside the cash. Apply to YC if your startup is generalist, B2B, deep-tech, or anything where the entertainment frame is a forced fit. Apply to both if you are on-thesis for Speedrun, but accept that the two programs run concurrent batches and you cannot do both at once.

The one underappreciated comparison: Sequoia Arc, which runs as a bi-annual open call paired with structured cloud and AI credits (up to $500K NVIDIA savings, $350K Azure credits, $250K Cloudflare credits, $100K Datadog) (Sequoia , Arc). For founders who could fit either Arc or Speedrun, the deciding factor is brand axis: a16z for entertainment-and-creative, Sequoia for enterprise-and-infrastructure.

What Speedrun acceptance actually requires

The sub-0.4% Speedrun acceptance bar is the public stat (a16z Speedrun , FAQ). What the FAQ does not say is what two human partners look for in five to ten minutes per application.

The qualitative signals, drawn from the structure of the questions and public partner write-ups:

• A non-obvious wedge: something the partner does not expect, in the first sentence. "We are building creator tooling" is obvious. "We are turning Steam workshop mods into a marketplace for studios that cannot afford a community manager" is not.
• A real why-now: a named technical or behavioural shift, dated, with a number where possible. Vague macro framing is the fastest path to skip.
• Founder-market fit: one sentence that makes the team feel inevitable for this product. Operator backgrounds, prior projects, a specific edge.
• Evidence of motion: any number that proves you are already executing. Users, revenue, partner conversations, design partners signed.
• On-thesis fit: explicit mapping to one of the thesis buckets. Implicit does not count.

In a sub-0.4% acceptance pool, the differentiator is rarely a brilliant idea. It is the specificity of the first sentence and the credibility of the why-now. Generic openings get cut in the first minute; specific openings buy you ten.

If you cannot answer all five qualitative signals truthfully in your application, you are competing for the small margin of submissions where partners take a flyer. That margin exists, but you are not the median accepted founder.

Who should not apply to a16z Speedrun

A counterintuitive section, because most accelerator guides tell every founder to apply. The bar is high enough that mis-fit applications waste the founder's time and the partners' attention.

Skip Speedrun if:

• You are post-Series A or at $1M+ ARR: you are past the Speedrun stage and the $500K-for-10% deal will dilute you for no benefit. Raise a priced round instead.
• You cannot relocate to San Francisco for 12 weeks: the program is in-person (a16z Speedrun , FAQ). Remote-only is not an option.
• Your startup has no entertainment, AI, or creative-tools angle: the thesis is narrow enough that pure-B2B-SaaS applicants are wasting cycles. Apply to YC, Sequoia Arc, or a sector-fit accelerator instead.
• You are based in a jurisdiction Speedrun does not invest in: international eligibility is open to founders from 30+ countries, but the list is finite. Check the home-page list before applying (a16z Speedrun , Apply).
• You are pre-product with no team: solo founders are accepted, but only with a co-founder or equivalent building capability on the team (a16z Speedrun , FAQ). "I have an idea" is not enough.

If Speedrun is one track of a parallel raising process and you are running outbound investor outreach alongside it, tools like Causo handle the VC matching and outreach so the Speedrun decision does not gate the rest of your round.

FAQ

What is a16z Speedrun? a16z Speedrun is Andreessen Horowitz's in-house accelerator for tech-meets-entertainment startups. Since launching in 2023 it has deployed over $300M to 250+ companies and runs as a 12-week San Francisco cohort twice a year. The program funds AI, 3D, creative tools, gaming, streaming and XR.

How much does a16z Speedrun invest? Speedrun invests up to $1M per accepted company. The upfront tranche is $500K for 10% common via SAFE; the second $500K goes into your next round within 18 months on the same terms as your other lead investors. The deal is the same for every accepted company.

What is the Speedrun application like? The application is a short text form with five questions: what you are building, who it is for, why now, why you, and what is working. Two human partners review every submission with no automated screening. There is no deck upload required, though a brief video may be requested at the next stage.

Is Speedrun only for games and AI? No. Since the March 2025 rebrand, Speedrun's thesis is tech-meets-entertainment, which covers AI, 3D, creative tools, gaming, streaming and XR. Pure-B2B SaaS startups with no entertainment or consumer angle are still a poor fit, but games-only positioning is outdated.

Does a16z Speedrun take a board seat? No. a16z explicitly does not take a board seat at the Speedrun stage, on the stated principle that founders should control their boards at this stage. You retain board control through the program and into the next round.

Related on the hub

• How to apply to 500 Global in 2026 — Related accelerators guide.
• How to apply to Sequoia Arc in 2026 — Related accelerators guide.
• How to apply to Y Combinator in 2026 (deadline playbook) — Related accelerators guide.