The H1 2026 State of Seed Fundraising Report

The state of seed fundraising in H1 2026 is bifurcated. Median seed pre-money sits at $18.4M and median round size at $3.0M, with seed-stage capital up 31% year-over-year to $12B in Q1 2026. But pre-seed instrument count fell 13% in 2025: fewer rounds, larger checks, AI taking the lion's share.

• Seed deal volume in 2026: more dollars, fewer rounds
• The seed market 2026 by the numbers
• Seed funding trends 2026: the AI bifurcation
• How long does a seed raise take in 2026
• What's actually clearing the bar versus stalling
• Dilution and equity sold at seed in 2026
• What H2 2026 looks like for seed founders
• How to read the rest of this state of seed report
• FAQ

Pull the H1 2026 state of seed fundraising data together from PitchBook-NVCA, Carta, and OpenVC and the picture is not "venture is back." It is "venture is concentrated." Total Q1 2026 venture deal value topped every prior full-year quarter at $267.2B per PitchBook-NVCA, yet pre-seed instrument count contracted 13% in 2025 per Carta's State of Pre-Seed 2025, and the survivors closed bigger. This report is the synthesized quarterly read of the seed market 2026: deal volume, round-size and valuation movement, time-to-close, and what is actually clearing the bar versus what is stalling. Every figure has a primary-source citation, and every benchmark links to the evergreen page where you can drill deeper.

Seed deal volume in 2026: more dollars, fewer rounds

Seed deal volume in 2026 is up in dollars and down in count. Capital is concentrating in fewer survivors at larger check sizes, with AI absorbing the dominant share. The headline numbers, primary sources only:

The headline: the median Q1 2026 seed round is $3.0M raised at an $18.4M pre-money valuation, per PitchBook-NVCA. That figure is "more than double the figure of 2021," per the same report. The subhead: Carta's pre-seed cohort shrank 13% in instrument count year-over-year in 2025 while total cash invested fell only 1%. Survivors raised more.

Stop reading the SERP headlines that say "venture is back" and read the deal-count line instead. Aggregate dollars going up while instrument counts drop is the signature of a concentration regime, not a recovery.

The seed market 2026 by the numbers

Reconciling PitchBook and Carta is the first thing every founder gets wrong. They publish different numbers because they measure different things. Here is the clean read.

PitchBook-NVCA tracks priced rounds across the full US venture market. Their Q1 2026 Venture Monitor reports median seed pre-money at $18.4M (AI $18.7M, Non-AI $18.0M) and median seed deal value at $3.0M. PitchBook's seed numbers were $16.0M pre-money and $3.8M deal value in their Q4 2025 Venture Monitor, so quarter-over-quarter the pre-money is up roughly 15% and the round size is moderately down.

Carta measures its own customer base, which skews toward later, more US-coastal, more AI-heavy companies. Their Record-Setting Early Stage Valuations post puts median seed post-money at $24M for Q4 2025, up from $18M Q4 2024 and $16M Q4 2023, a 50% two-year increase. Their State of Pre-Seed Q1 2026 tracks 3,000 US-based Carta startups raising $2.3B+ in pre-seed funding in Q1 2026, holding flat with recent quarters.

Which number do you use to set your ask? Use PitchBook's median pre-money for your valuation comparable, because it reflects the broader market and is not skewed by Carta's later-stage customer concentration. Use Carta's post-money trend line to understand directional momentum at the higher end of the market. Both signal the same thing: 2026 is the most expensive seed cohort on record.

For an explicit valuation read, the seed round valuation benchmark 2026 page breaks the same data out by sector and check size.

Seed funding trends 2026: the AI bifurcation

More than 60 cents of every venture dollar in Q1 2026 went to AI, per Carta's State of Private Markets Q1 2026. That single fact distorts every aggregated seed benchmark on the SERP, and explains why the founder you talked to last week at a YC event quoted a $25M pre-money like it was normal.

Peter Walker, Head of Insights at Carta, breaks the 2026 seed market into three tiers on the PMF Show:

• AI application seed rounds: $4-5M raised on $20-22M post-money. This is the new "normal" seed for an AI-native company shipping product to a real customer segment.
• AI infrastructure seed rounds: priced on $160-200M post-money, which Walker characterizes as "astronomical numbers." These are model-layer, compute-layer, or core-tooling bets where the round is a strategic option, not a valuation in the traditional sense.
• Non-AI seed: tracks the PitchBook non-AI median of $18.0M pre-money on roughly $3.0M raised. This is the legacy benchmark, increasingly disconnected from the headline numbers.

Sequoia frames the macro context in Building Tomorrow's Transformational Companies as "a foundational platform shift as transformative as the rise of the internet," with the caveat that "with any tech supercycle, there will be both euphoria and troughs of disillusionment." Translation: the AI infra tier is priced on optionality, the AI app tier is priced on near-term traction, and non-AI is priced on the old rules.

The implication for your own raise: if you are not AI-adjacent, do not index on the $24M Carta median post-money when you set your ask. That figure is pulled upward by AI infra outliers and by Carta's customer mix. Your relevant comparable is the non-AI PitchBook median, which sits much closer to a $3M raise at an $18M pre-money. The seed valuation 2026 deep dive splits the data by AI vs non-AI, by sector, and by region.

How long does a seed raise take in 2026

Plan for three to six months from kickoff to wire, with American companies trending faster than international ones. Per CRV's what seed investors look for in 2026 post, raising seed funding "takes most founders three to six months," with American companies typically three to four months and international founders four to six.

That is the headline. The ranges hide three reality checks.

Faster closes pair a warm intro with an AI wedge. A founder who shows up with an introduction from a portfolio CEO, a defensible AI angle, and clear early traction is closing on a four-week sprint. A founder cold-emailing into associates without an intro or a wedge takes the full six months and often comes back empty.

The Series A bar is dragging the seed bar up. Peter Walker states on the PMF Show that "the bar for Series A is something like 2x what it was a couple of years ago." That tightening flows backward: seed investors are now underwriting whether you can clear the harder Series A in 18 months, not just whether your prototype works.

Geography matters more than it used to. US founders raising from US funds close at the 3-4 month end. European, LatAm, and APAC founders raising from US funds add a discovery and diligence overhead that pushes them to 4-6 months. UK and EU funds are faster on their own continents but slower to deploy on AI-native seeds outside their geographic mandate.

Map the calendar before you start. The how much to raise at seed guide walks through the round-sizing math; combine it with a realistic time-to-close to set runway requirements.

What's actually clearing the bar versus stalling

The Q1 2026 seed market is not friendly to default-dead pitches. Kruze's March 2026 seed guide declares "2026 is the year that discipline and concentration define who actually gets funded," characterized by "fewer seed rounds, larger checks" versus 2025, with AI-native startups capturing disproportionate share. That matches the Carta deal-count contraction precisely.

What is clearing the bar in H1 2026:

• AI-native companies with revenue. Any AI app showing real ARR, even at the $20-50K MRR range, prices on the $4-5M raise / $20-22M post-money curve Walker described, fast.
• Repeat founders with an AI wedge. Even with no revenue, a credible operator + AI angle clears at the same level. Investor underwriting weight shifts heavily toward execution risk in this cohort.
• Infrastructure plays adjacent to AI demand. Vector databases, eval frameworks, agent harnesses, data pipelines for training data. Priced like the AI app tier or higher.
• Vertical AI for regulated markets. Health, fintech, legal AI where domain insight is the moat. Investor appetite is selective but pre-money has lifted to the AI app tier when traction is real.

What is stalling:

• Non-AI consumer. Outside of breakout viral metrics, hard to clear.
• Solo founder, first-time, no AI angle. Default-dead pitches are largely unsuccessful per the PMF Show analysis, "absent an exceptional growth story."
• Pre-revenue SaaS pitches with a $3M ask at a $20M pre-money. Investor pattern recognition rejects the math when there is no AI wedge to justify the premium.
• International seeds raising into US funds without a US entity. The Delaware C-corp gate remains the structural filter; OpenVC's 2026 guide characterizes the year as "bigger but harder" and warns that "supergiant seed rounds just provide more capital to burn through as they fall."

The 2026 seed market is not a 31% expansion, it is a 13% deal-count contraction masked by larger checks concentrated in AI.

The contraction is the story. The expansion in dollars is a side effect of where the dollars are clustering. If your pitch does not fit the cluster, the median benchmark numbers do not apply to you. The seed-to-Series A graduation rate benchmarks 2026 page tracks what happens to companies on either side of the bar 18 months out.

Dilution and equity sold at seed in 2026

Seed investors typically take 15 to 20 percent equity in a seed round per CRV. That range is the durable structural number across the cycle; what has changed in 2026 is the absolute check size that buys that share.

Working the math: a $3.0M round at an $18.4M pre-money sells roughly 14.0% in a clean priced round (3.0 / 21.4). A $4.5M AI app round at a $20M pre-money sells 18.4%. Both land inside the CRV range, but the dollar amount and the implied valuation are very different.

SAFEs distort the picture. Carta reports 50,316 SAFEs and convertible notes issued in 2025 totaling $10.4B, with median post-money SAFE caps around $10M on $250K-$1M rounds and $15M on $1M-$2.5M rounds. The SAFE cap is the dilution gating mechanism for most pre-seed and small seed rounds, not the priced pre-money.

Do not stack SAFEs to hide cumulative dilution. Two $750K SAFEs on a $10M cap and a $15M cap, followed by a $2M priced round at $18M pre, can quietly add up to a 25%+ dilution before the priced round even starts. The cap table model needs to convert every note explicitly. Carta and Pulley both have free tools for this; do not run the model on a back-of-envelope.

Practical implications:

• If you raise a single $3M priced round at $18M pre, plan for 14-15% dilution. This is the cleanest seed structure available.
• If you raise on stacked SAFEs, plan for 18-25%+ cumulative dilution by the time the cap converts. Each note carries optionality cost.
• If a fund offers 25%+ equity at seed, that is a structural mismatch. US seed funds price for 15-20%; a 25%+ ask usually signals an accelerator-adjacent term sheet or a misread of the round size.

What H2 2026 looks like for seed founders

Expect the bifurcation to widen, not narrow. Sequoia's read of the AI supercycle in Building Tomorrow's Transformational Companies sets the macro pattern: euphoric capital deployment at the top, troughs of disillusionment for the median company, both happening in parallel.

The H2 2026 directional calls from the H1 data:

• Median pre-money likely stays in the $18-20M zone for non-AI, given the PitchBook trajectory and the absence of any signal that Series A is loosening. The PMF Show framing of "Series A bar at 2x" is the gravitational pull keeping the seed bar high.
• AI app tier likely stabilizes around $20-25M post-money, with AI infra continuing to price as strategic options at $100M+ post. The Walker numbers from May 2026 will be the reference benchmarks through year-end barring a macro shock.
• Deal count continues to contract in non-AI segments if the Carta 2025 trend extends. The $12B Q1 2026 OpenVC figure (+31% YoY) is the dollar headline; the deal-count line is what actually tells you whether the funnel is open for your pitch.
• Time-to-close stays at 3-6 months for the median raise, with founders who cannot hit the AI bar drifting toward the six-month end. Set runway expectations on the slower side of the range if your wedge is not AI-native.

The actionable read: if you are starting a raise now, plan for a six-month process, a single priced round structure to avoid SAFE stacking, and a $3-4M ask sized to 18 months of runway at the new Series A bar. Anything else is fighting the H1 2026 distribution.

How to read the rest of this state of seed report

This page is the quarterly narrative anchor. Drill into the specific number you need from the evergreen benchmark pages:

• For the explicit pre-money and post-money split by sector, AI vs non-AI, geography, and check size, use the seed round valuation benchmark 2026.
• For the deeper read on what the headline median actually means for your specific company stage and traction profile, the seed valuation 2026 deep dive is the right anchor.
• For round-sizing math (how much to ask, how to convert ask into runway, how to size against your dilution tolerance), use how much to raise at seed.
• For what happens to your company on the other side of the round, the seed-to-Series A graduation rate benchmarks 2026 page tracks the conversion data the 2x Series A bar implies for your eventual Series A odds.

If you are running cold outreach as part of the raise, tools like Causo handle the personalization at scale so you can hit the higher message volume the 2026 funnel demands without losing reply rate.

FAQ

What is happening in seed fundraising in 2026? Seed activity in H1 2026 is stable in deal count but concentrated in AI and in larger checks. Median seed pre-money valuation reached $18.4M in Q1 2026 per PitchBook-NVCA, up from $16M in Q4 2025, while pre-seed instrument count contracted 13% year-over-year in 2025 per Carta. Fewer rounds, bigger checks, and AI capturing more than 60 cents of every venture dollar.

How much are seed rounds in 2026? Median seed round size is $3.0M in Q1 2026 per PitchBook-NVCA, with median seed pre-money valuation at $18.4M and median post-money around $24M per Carta's Q4 2025 read. AI application seed rounds typically close at $4-5M on $20-22M post-money, while AI infrastructure seed rounds can price on $160-200M post per Peter Walker, Carta's Head of Insights.

Is seed funding up or down in 2026? Dollars are up, deal count is down. OpenVC reports $12B aggregated into seed-stage capital in Q1 2026, a 31% year-over-year increase. Carta reports a 13% drop in pre-seed instrument count over 2025, meaning the median check is larger while fewer companies are raising at all.

How long does a seed raise take in 2026? Most founders take three to six months to close a seed round per CRV, with American companies trending three to four months and international founders four to six. Faster closes usually pair a warm-intro path with a clear AI wedge and traction that clears a recently raised Series A bar.

What's the average dilution in a seed round in 2026? Seed investors typically take 15 to 20 percent equity in a seed round per CRV. At a $3.0M round on an $18.4M pre-money, you sell roughly 14% in a priced round. SAFE caps run lower at the smaller end, with median post-money caps around $10M on $250K-$1M raises per Carta, so cumulative dilution depends on how many notes stack.

Related on the hub

• Seed fintech VCs US: the 2026 velocity-ranked list — US fintech seed VCs ranked by deployment velocity in the last 90 days. Skip funds that drifted to Se…
• The H1 2026 AI startup funding report — Related fundraising basics guide.
• The H1 2026 Pre-Seed Funding Report — Related fundraising basics guide.
• Seed round valuation 2026: the benchmark report — Related fundraising basics guide.