How to price an AI product before Series A.

Before Series A, prepare a price, cost and margin plan you can explain in diligence.

Before the round, connect pricing to the costs that vary with usage, instrument gross margin at the account level and test retention. A hybrid can help with variable costs, but the rate still has to cover delivery. These are diligence inputs, not a formula that guarantees a raise. Investors can choose thin margins in exchange for growth; you still need to explain the tradeoff and how you will fund it.

Megadeals make poor pricing benchmarks

The final Q1 2026 PitchBook-NVCA Venture Monitor reports $267.2B in US venture deal value across sectors. Five deals account for $195.6B, or 73.2 percent of it. This is a US all-sector dataset, not global AI funding. Its concentration explains why a headline total tells an application founder little about an achievable round.

An all-stage average also mixes companies with different capital needs and traction. Choose comparisons for your stage and product category, then build your pricing around customer value and delivery costs. A frontier lab's financing does not tell you what to charge for an application.

Read the Series A benchmark with its cohort

In his May 25, 2026 Product Market Fit Show interview, Carta's Peter Walker described median Series A raises around $13M to $15M and post-money valuations around $75M to $85M. For revenue, use the source chart: SVB's H1 2026 report, page 28 gives a $2.5M median trailing-four-quarter total revenue for US technology companies raising Series A in pooled 2024 to 2025, excluding extensions. It includes recurring and one-time revenue, not ARR, and is not AI-only. Page 29 separately reports a 2025 median of $3.3M versus $1.6M in 2021. Neither chart sets a minimum revenue requirement for a 2026 AI round.

In the same interview, Walker described elite AI infrastructure and foundational-model seed rounds at $160M to $200M post-money. Those valuations concern a narrow cohort. Application founders need comparisons reflecting their own market and evidence of demand, retention and cost to serve.

Walker also described median seed headcount of four employees, excluding founders, and Series A headcount of roughly fifteen to seventeen. Those are dated interview estimates, not an AI-only staffing requirement. With a team that size, a founder often has to own pricing before there is a dedicated hire.

Two-tier valuations: why your comps are probably wrong

Carta's Q1 2026 report distinguishes foundational-model Series A valuations around $300M from non-AI valuations around $55M, calling them different markets. Its 2025 AI fundraising data puts the overall AI Series A valuation premium at 38 percent. That broader AI category is not the same as foundational models, and neither is a quote for your company.

An application built on someone else's model should use relevant application comparisons. Investors still have to assess its data, workflow, distribution and economics. The AI label alone does not establish a premium; a credible customer and delivery-cost story gives you something specific to discuss.

What "price for margin before Series A" actually means

Start with the variable compute cost behind the price. A heavy user on a flat plan can produce a growing serving bill without bringing more revenue. Compare what that account pays with its delivery costs before describing usage growth as healthy demand. Diligence needs the account evidence and the plan for any loss you intend to fund.

For variable-cost products, I recommend a platform fee for access and the workflow or support around the model, plus a usage component for expensive consumption. Set the included allowance and overage rate against measured account costs. Stable, low-cost usage can support seats; unlimited consumption needs a cost test before you promise it.

I co-founded the AI fintech FaaStrak, which went from zero to $1M ARR in nine months. For the implementation mechanics, the separate electrical-distribution AI SaaS case describes usage pricing and per-customer model-cost metering. The account-level metering playbook shows the tagging and rollup needed to prepare cost-to-serve evidence before diligence.

At pre-seed to Series A, the founder often owns the commercial AI decisions: pricing, unit economics, customer demand and build-or-buy choices. Assign those decisions now, whether they stay with you or go to another owner.

Fast revenue is not the same as fundable revenue

A demo tests capability. Diligence also tests the revenue records, customer retention and costs behind it. Keep ARR, booked revenue and total revenue separately labelled, and make the account-level costs traceable to charges. A usage spike alone does not answer whether customers will renew at a price you can support.

Carta reports an 11.4 percent down-round rate in Q1 2026. That market-wide figure does not prove that margin-positive AI companies alone get funded. Your task is narrower: show which customers stay, what they pay and what it costs to deliver the work, including any planned period of low margin.

What to do before your Series A

If you do nothing else this quarter, do these, in this order:

Pick relevant comparisons. Use companies with a similar product category, stage and delivery model. Foundational-model and infrastructure financing may tell an application company little about its own valuation or revenue expectations.

Connect price to cost. Test a tenfold usage increase against the current plan. If its cost exceeds the revenue available to cover it, revise the allowance, tier or usage rate before the round.

Instrument margin per account. Know your gross margin and your cost-to-serve at the account level, not just in aggregate. Show serving costs and the price against them, including your allocation assumptions.

Show retention alongside growth. Track cohorts that renew and expand, and explain any usage spike followed by churn.

Label revenue precisely. SVB's $2.5M median is trailing-four-quarter total revenue for US technology Series A raises in pooled 2024 to 2025, excluding extensions. Show your own recurring and one-time revenue separately, plus growth, retention and margin. The median is context, not a cutoff.

Common questions on pricing before a Series A

Do you need revenue to raise a Series A for an AI startup? Revenue and retention give investors evidence to assess, but pre-revenue rounds exist and no single cutoff applies to every company. Use comparisons for your stage and product category. Foundational-model financing and an application company's raise can have very different requirements.

How much revenue do you need for an AI Series A in 2026? There is no verified AI-only minimum here. SVB's H1 2026 report gives a $2.5M median trailing-four-quarter total revenue for US technology companies raising Series A in pooled 2024 to 2025, excluding extensions. Its separate 2025 chart gives $3.3M. Both include recurring and one-time revenue, not ARR. Investors assess the company's growth, retention, costs and market alongside revenue.

Should AI products use usage-based or seat-based pricing? A hybrid can suit variable serving costs: a base fee with included usage, then a usage charge for expensive consumption. Test the allowance and rate against account costs. Pure seats can work when usage is stable and compute is a small, predictable share of the price. Usage pricing also needs a suitable unit and a rate that covers delivery.

What valuation can an AI application company expect at Series A? There is no single application-company valuation implied by these figures. Carta reports a 38 percent overall AI Series A valuation premium for 2025 and separates foundational-model companies in its Q1 2026 report. Use relevant application comparisons and show evidence of customer demand, retention and economics rather than treating a frontier-model valuation as your benchmark.

Before diligence, bring account costs, pricing allowances and rates, retention cohorts and a clear margin plan. Show how a heavy user changes both your revenue and your serving bill. That is the evidence needed to discuss pricing and growth without guessing.

If you are staring at this gap and the next round is closer than the answer, that is the conversation I have.