The H1 2026 Series A Bar Report

The Series A bar in H1 2026 is roughly $3M ARR, 2x YoY growth, burn multiple under 1, and strong cohort expansion, clearing at a $48M median pre-money for non-AI companies. AI foundational startups raise at a $300M median, 5.4x non-AI peers. Only ~20% of seed companies make it, and the median seed-to-A wait is now 616 days.

• The 2026 Series A bar in one table
• ARR: where Series A requirements actually sit
• The seed to Series A 2026 graduation rate
• AI vs non-AI: two different Series A benchmarks
• Time from seed to A: the 616-day reality
• Burn multiple, NRR, and the efficiency cut
• Valuation: what a clean Series A term sheet looks like in 2026
• What to do this quarter

There is no single Series A bar in 2026. There are two, and they are 5.4x apart.

If you are running an AI foundational model company, the median pre-money you will clear at is $300M. If you are running anything else, the same median is $55M, per Carta State of Private Markets Q1 2026. The metrics, the timeline, and the dilution math all bifurcate from there.

This guide is the operator's read of where the Series A bar actually sits in mid-2026: the ARR, growth, NRR, burn multiple, and valuation thresholds that close term sheets right now. It is also a warning. Median deal counts are collapsing, the graduation rate is roughly 20%, and the time from seed to A has stretched to a record 616 days. The bar is higher and the wait is longer.

The 2026 Series A bar in one table

The compressed view. Everything below is elaboration.

Read this table as your scorecard. If you are missing two or more rows for your stage, you are not at the bar yet, and you should plan an extension rather than push into a process you will lose.

ARR: where Series A requirements actually sit

$3M ARR is the new floor for a clean Series A in non-AI SaaS. That is the number Peter Walker repeated on the Carta Q2 2025 State of Private Markets podcast, and it is the line that mid-2026 partner pitches converge on.

The OpenVC founder-facing read is more lenient: at least $1M ARR plus 2x YoY growth as the minimum to enter a process, per OpenVC's 2026 valuation guide. The gap between those two numbers, $1M floor versus $3M median, is the gap between a hard process and a competitive one. At $1M ARR you can get meetings if your growth is loud enough. At $3M ARR with a burn multiple under 1, partners come to you.

Do not optimize for the floor. Founders who target the bottom of the range raise on bad terms or do not raise at all. The graduation rate of ~20% (covered below) means the typical seed company is closer to dead than to a Series A, and the bar that gets you a competitive process is consistently higher than the bar that gets you a term sheet at all.

✅ Good: "We are at $3.2M ARR, 2.4x YoY, burn multiple 0.8, NRR 128%. Closing this quarter." It reads as a real Series A pitch because every number is at or above the median, and the founder is committing to a close. ❌ Bad: "We are at $1.1M ARR, 'pacing' to $2M. Open to chatting whenever." It reads as an extension pitch dressed as a Series A, because pacing claims are trailing numbers dressed in forward-looking clothes.

Calibrate which one you are running and pitch it honestly. For the seed-stage benchmarks that feed this calculation, see seed MRR benchmarks for 2026.

The seed to Series A 2026 graduation rate

Only about 20% of seed-funded startups graduate to a Series A. That figure comes from the Carta Q2 2025 State of Private Markets podcast and it is the single most important number in this guide.

Five out of six seed companies do not raise a Series A inside the standard window. They run out of runway, raise a bridge, get acqui-hired, shut down, or stay in seed-extension land for years. The 20% graduation rate 2026 view is not a quirk of a tough market. It is the structural reality of the seed to Series A 2026 path.

That changes how you plan. Your seed plan is not a Series A plan with the dates moved up. It is a plan to get into the top quintile of your cohort, or to extend long enough that you can. The Rob Hayes minimum, per First Round Review, is to build runway assuming zero revenue, exclude venture debt from the runway calc, and spend 50%+ of your time hiring. Founders who follow that math survive long enough to be one of the 20%.

Concentration makes the cut sharper. Roughly half of all US VC deal value in 2025 flowed into just 0.05% of completed deals, per PitchBook-NVCA Q4 2025. The median Series A bar is rising because the top of the market is sucking up the capital, leaving thinner air for everyone in the middle. For deeper coverage of the funnel math, see seed to Series A graduation rate benchmarks for 2026.

AI vs non-AI: two different Series A benchmarks

The single biggest factor in your 2026 Series A benchmarks is whether you are positioned as AI.

The median pre-money for an AI foundational Series A in Q1 2026 was $300M. For non-AI it was $55M. That is a 5.4x gap on the same nominal stage, per Carta Q1 2026. On valuation alone, "Series A" means two different things now.

The capital concentration is even more extreme than the valuation gap suggests. AI startups took 65.4% of all US venture deal value in 2025, up 16.3% year over year, and 39.4% of deal count, per PitchBook-NVCA Q4 2025. When over 60% of dollars chase one category, the bar for that category gets re-priced. The bar for everything else gets thinner air.

The Series A AI premium is not free money. It comes with a higher metrics bar, not a lower one. The a16z 2025 thesis, per Big Ideas in Tech 2025, is that AI-native winners at Series A are graduating from "point tools" into "core systems of record." A point-tool AI wrapper raising at the AI median is the most over-funded position on the field, and it tends to flame out at the B. A vertical workflow owner raising at the same median is the most under-funded position, because the long-term math says they own a category.

Position accordingly. If you are AI-adjacent, get the AI label honestly. If you are not AI, do not fake it: pretending to be AI at Series A diligence is the fastest way to lose the round mid-process. Honest positioning beats fashionable positioning every time, because the diligence team checks.

Time from seed to A: the 616-day reality

The median seed-to-A interval is now 616 days, or roughly 20 months. That is the highest figure on record in the PitchBook-NVCA Venture Monitor series, per Q4 2025. The "12 to 18 month seed runway" advice that founders hear from accelerators is now structurally short of reality.

What this means for your plan: build the runway for 24 months, not 18. If you raise a seed in January 2026 with 18 months of runway, the median path puts your Series A two months after your money runs out. That is not a margin of safety. That is a forced bridge.

Two ways to handle it:

• Raise more at seed. A $4M seed at $25M post-money is much easier to extend than a $2M seed at $12M post. If you can justify the bigger seed, take it. The dilution looks worse on paper and better in cash.
• Plan the seed extension upfront. Treat a 6-month extension at month 14 as the base case, not the disaster case. Founders who plan extensions in advance raise them at flat or slightly-up terms. Founders who run out of cash first raise them at down rounds with structure.

For deeper coverage of when an extension is the right call versus pushing into a Series A process, see bridge vs Series A timing signals.

Burn multiple, NRR, and the efficiency cut

The 2026 efficiency bar is harder than 2021 but easier than 2023. The number to know: burn multiple under 1.0 is the Series A bar. That is the operating reality that CRV's March 2026 metrics piece and Waveup's April 2026 fundraising guide converge on, and it is what term sheets in the wild reflect.

Burn multiple is net burn divided by net new ARR. A burn multiple of 1.0 means you burned $1 to add $1 of ARR. Under 1.0 means you are getting more ARR than the dollar you spent. Above 2.0 is essentially a fail, except for AI companies with a story. Above 3.0 is a fail with no exception.

NRR (Net Revenue Retention) is the second efficiency cut, and the one founders consistently underestimate. 100% NRR is the floor; 120%+ is the bar for a competitive process. A 95% NRR with a great new-logo story still loses the round to a 120% NRR with a mediocre new-logo story. Existing-customer expansion is what convinces a Series A partner you have product-market fit.

Do not chase the premium if you do not have it. Raise on the metric you actually have, not on the one you are pacing to. Pacing claims are the single biggest reason term sheets fall through in diligence, because the diligence team prices to the trailing number, not the projected one. For benchmarks on what your current burn multiple says about your timing, see burn multiple benchmarks for seed startups in 2026.

The Series A AI premium is no longer a rumor: median pre-money is $300M for AI foundational startups versus $55M for everyone else, a 5.4x gap on the same nominal stage.

Valuation: what a clean Series A term sheet looks like in 2026

The headline Series A metrics 2026 for valuation, in order:

• Median post-money hit $78.7M in Q4 2025, up 37% YoY, per Carta Record-Setting Early-Stage Valuations.
• Median pre-money hit an all-time Carta high of $49.3M in Q3 2025, per Carta Q3 2025.
• Median deal count fell 18% YoY in Q2 2025 with cash raised down 23% to $4.7B, even as primary Series A valuations rose 20% YoY to a $47.9M median, per Carta Q2 2025.
• Q1 2025 clean baseline: $7.9M median round at $48M median valuation, per Carta Learn.
• Kruze sample: US Series A median pre-money rose from $20.3M in 2024 to $25.3M in 2025, per Kruze Consulting, tracking a smaller-sample view of the same upward trend.

The contrarian point: valuation is the least important number on your term sheet. The structure matters more than the headline.

YC's definition of a clean Series A term sheet, per Y Combinator Startup Library, excludes:

• Liquidation preferences above 1x. A 1.5x or 2x preference looks small until the exit math runs and your common stock gets wiped.
• Participating preferred. This means the preferred shareholder gets their money back AND a pro-rata share of the rest, which is double-dipping at exit.
• Cumulative dividends. Dividends that accrue and stack create a hidden preference that compounds over time.
• Warrant coverage. Warrants are free options for the investor that dilute you at exercise without adding capital.
• A 2-2-1 board. YC's view is founders should retain a 2-1 board to avoid deadlocks. Two-two-one boards turn every disagreement into a fundraising negotiation.

Take the lower valuation with clean terms. A $40M pre-money with a 1x non-participating preferred is dramatically better than a $60M pre-money with a 1.5x participating preferred and 20% warrant coverage. The dirty deal looks like a higher valuation. The cap-table math says otherwise, and dirty terms are nearly impossible to clean up in later rounds.

The dilution context matters too: founders own just 56% of their company after the first priced round, and roughly 10% by Series D, per the Carta Q2 2025 podcast. Every percentage point you give up on dirty terms compounds across the rest of the cap table.

Pre-stage your data room before signing. The YC Series A diligence checklist covers seven categories: corporate records, business plan and financials, IP assignments, security issuances, material agreements above $25K, ongoing disputes, and employee benefits. Founders who pre-stage close two to four weeks faster, which is often the difference between closing in this market and watching it close on you.

What to do this quarter

The action list.

1. Score yourself against the table. Use the H1 2026 Series A bar table above. Count the rows where you are at floor, at bar, and at premium. Two or more "below floor" rows means start the seed-extension conversation, not the Series A conversation.
2. Get the NRR number right. Most seed-stage founders do not track NRR cleanly because they do not have enough cohort history. Build the cohort table in your data room first. A 120% NRR with three cohorts of history is more credible than a 140% NRR computed off one quarter.
3. Pre-stage the diligence list. Run the YC seven-category checklist before you take any partner meetings. Two of those categories (IP assignments and material agreements above $25K) routinely tank Series A processes that look clean from the outside.
4. Decide the AI framing. Are you AI-foundational, AI-applied, or non-AI? Pick one honestly and pitch only that one. The Series A diligence team checks, and partners burn the founder who repositioned themselves as AI for the round.
5. Build the runway for 24 months. The 616-day median seed-to-A interval means 18-month seed runway is now under-planned. Either raise more at seed or plan a 6-month extension at month 14 as your base case.
6. Run an outreach list of 40 to 80 partners. At Series A, partner-level fit is everything. Hit only partners with explicit thesis fit and a recent investment that maps to your space. For the timing and personalization workload at this volume, tools like Causo automate the mechanics so you can spend the cycles on the pitch.

The Series A bar in 2026 is not impossible. It is narrower and more bifurcated than it was, and the founders who clear it are the ones who score themselves honestly against the table, pick the right framing, and run a structured process. Everything else is a bridge in disguise.

FAQ

What is the minimum ARR needed to raise a Series A in 2026? The de-facto minimum for non-AI B2B SaaS is $1M ARR with 2x YoY growth, and the median is $3M ARR. AI-foundational companies can raise at lower ARR if the model and team are strong, but the metrics bar is still real. Below $1M ARR you are pitching a seed extension, not a Series A.

What is the seed-to-Series A graduation rate (what % of seed startups make it)? Roughly 20% of seed-funded startups graduate to a Series A, per the Carta Q2 2025 State of Private Markets podcast. The other 80% bridge, extend, get acqui-hired, or shut down. The graduation rate is the structural reality of seed, and your runway should be planned around it.

Do AI startups face different Series A benchmarks than traditional SaaS companies? Yes, and the gap is enormous. AI foundational Series A median pre-money was $300M in Q1 2026 versus $55M for non-AI, per Carta. The growth bar is higher (3x+ versus 2x), but ARR thresholds are softer for AI startups with strong models and team signals.

How long from seed to Series A? The median seed-to-next-priced-round interval reached 616 days, roughly 20 months, in Q2 2025, the longest stretch on record, per PitchBook-NVCA Q4 2025. The old 12-to-18-month rule is now under-planned. Build seed runway for 24 months or expect to raise a bridge.

What's a good burn multiple for a Series A? Under 1.0 is the bar. Above 2.0 is a fail except for AI companies with a strong narrative; above 3.0 is a fail in any scenario. Burn multiple is calculated as net burn divided by net new ARR, and Series A partners read it as the proof of efficient scaling.

Related on the hub

• Series A climate VCs actively deploying in 2026 — Series A climate VCs writing checks in 2026, ranked by deployment velocity, with check sizes, sector…
• Series A AI VCs deploying in 2026: the global list — Global Series A AI VCs actively deploying in 2026, split by foundation-model-capable funds and appli…
• Series A data room checklist: 58 documents for 2026 — Series a data room checklist for 2026: 58 documents grouped by folder, with cohort, reference, and K…
• The H1 2026 Venture Market Report for Founders — Related vc process guide.
• The H1 2026 AI startup funding report — Related fundraising basics guide.
• The H1 2026 Pre-Seed Funding Report — Related fundraising basics guide.