Agentic AI Startup Funding 2025-2026

What are all the funding deals in the agentic AI market from August 2025 to September 2026?

The table below lists every disclosed equity financing in the dataset for pure-play agentic AI companies between August 2025 and September 2026. We define the agentic AI market as software that can turn a user goal into a multi-step plan and take actions through tools or integrations to produce an outcome.

We include agentic applications and platforms that manage task state through context or memory, execute and verify steps, and operate with either full autonomy or human approval gates. We exclude copilots that only suggest content, one-off tool calls without planning or state, and deterministic RPA, BPM or iPaaS automation where an LLM does not drive step selection or execution. For a broader view of the competitive landscape, see our Agentic AI market report.

How active has fundraising been in the agentic AI market?

As of September 2026, fundraising in the agentic AI market has been active but highly uneven over the requested 12 months. The dataset contains 39 disclosed equity financings across 32 unique companies, representing approximately $3.676B of capital.

Deal flow averages 2.79 financings per calendar month, with a median of 2. The difference is modest, which suggests financing activity itself is relatively distributed even though dollars are not.

Capital flow is much more volatile. The agentic AI market averages $262.5M raised per calendar month, compared with a median monthly total of $181.25M, while some months recorded little or no qualifying activity.

The strongest reading rule is therefore to separate transaction activity from capital activity. A month can contain several legitimate agentic AI deals without moving the market total much unless one of the larger scale companies also raises.

For a deeper view of current companies, categories and funding patterns, see our full Agentic AI market report.

How concentrated has fundraising been in the agentic AI market?

As of September 2026, fundraising in the agentic AI market is heavily concentrated at the top over the requested 12 months. The largest financing represents 14.96% of disclosed capital, the top 3 reach 35.37%, the top 5 reach 47.61%, and the top 10 absorb 69.11%.

The largest individual financing is Legora's $550M Series D, followed by Cognition's $400M Series C and Sierra's $350M late-stage round. A small number of application leaders therefore have an outsized effect on total market funding.

This concentration is important because company count and capital concentration tell different stories. Thirty-two companies received funding, but the majority of dollars still accumulated around a far smaller group of established winners.

Total capital should therefore not be interpreted as evenly available across the agentic AI market. Funding conditions for a new memory, approval or execution-layer startup are very different from those facing an established legal or customer-service agent company.

How much of the agentic AI funding signal is driven by outliers?

As of September 2026, most of the dollar signal in the agentic AI market is driven by large outlier financings over the requested 12 months. Sixteen of 39 deals were strictly above $50M, yet those rounds supplied $3.131B, or 85.2% of all disclosed capital.

Capital excluding rounds above $50M totals only $544.5M. Removing the megarounds therefore eliminates roughly five-sixths of the headline market total even though most companies remain in the sample.

The $94.24M average round is consequently a poor benchmark for a typical financing. The median is $43M, making the average 2.19 times the median and confirming substantial right-tail distortion.

Twelve rounds also exceeded $100M, representing 30.77% of all financings. Large checks are not a single-company anomaly in the agentic AI market, but they remain concentrated in selected categories and market leaders.

Is the agentic AI market broad with many targets, or narrow with few fundable companies?

As of September 2026, the agentic AI market has a reasonably broad company set but a narrow pool of companies absorbing very large checks over the requested 12 months. Thirty-nine financings correspond to only 32 unique companies because several leaders raised more than once.

Repeat financing makes raw deal count look broader than new-company formation really is. Legora raised three times in the period, while Harvey, Factory, Wonderful, Mem0 and Keycard also produced multiple qualifying financings.

The distinction matters because repeat raises are generally concentrated among companies demonstrating enterprise adoption or rapid product deployment. Deal count therefore overstates the number of independently validated new entrants.

A useful way to read the agentic AI market is to track both unique funded companies and repeat fundraising velocity. A company that returns quickly at a higher stage provides a stronger signal of investor re-underwriting than a single isolated financing.

Is agentic AI mostly an early-stage formation market or a late-stage scaling market?

As of September 2026, the agentic AI market behaves more like a scaling market than an early-stage formation market over the requested 12 months. Seed and Series A rounds account for 53.85% of deal count but only 17.05% of disclosed capital.

Early-stage financing totals $626.7M. By comparison, Series B, Series C, Series D+ and Growth Equity rounds account for approximately $3.028B, or 82.38% of disclosed capital.

Series D+ alone attracted $1.540B, equal to 41.90% of all capital, while Series C added another $738M. Investors are putting a disproportionate share of dollars behind companies that already show signs of category leadership.

The financing labels also compress unusually quickly. Wonderful moved from a $100M Series A to a $150M Series B within four months, while Factory progressed from a $50M Series B to a $150M Series C in roughly seven months.

For more context on which companies are already scaling beyond formation-stage financing, see our analysis of the Agentic AI market.

Which categories attract the most investor attention in agentic AI?

As of September 2026, Vertical AI Agents attract by far the most investor attention in the agentic AI market over the requested 12 months. The category produced 22 of 39 disclosed financings and raised $3.038B, equal to 82.64% of all capital.

Agent Execution Infrastructure ranks second by deal count with 5 financings, but those rounds raised only $187M. Agent Development Platforms recorded 4 deals and $319.5M, while Agentic AI Applications also produced 4 deals but only $100.5M.

Agent Memory Systems remain active enough to form a distinct category, with 3 financings, yet they raised just $26.9M combined. Human Approval Agents contributed only one $4M qualifying financing.

The market is therefore giving much more capital to companies that own an end workflow than to companies supplying individual architectural components. Legal, customer-service, coding and finance agents repeatedly reach larger financing stages because their output maps directly to measurable work.

We cover the category structure and competitive landscape in more detail in our Agentic AI category report.

Which categories attract disproportionately large checks in the agentic AI market?

As of September 2026, Vertical AI Agents are the only major category attracting disproportionately large checks relative to deal count over the requested 12 months. Their capital-share-to-deal-share ratio is 1.47x.

Vertical AI Agents capture 82.64% of capital from 56.41% of financings, with an average round of $138.07M and a median of $87.5M. That premium shows investors are underwriting workflow ownership more aggressively than horizontal tooling.

Agent Development Platforms are closer to parity at 0.85x, with a $79.88M average round and $87.5M median. Agent Execution Infrastructure falls to 0.40x despite representing 12.82% of all deals.

Agentic AI Applications register a 0.27x ratio, Agent Memory Systems 0.10x, and Human Approval Agents just 0.04x. Architectural importance therefore does not automatically translate into venture-market value.

Which geographies matter most for fundraising in the agentic AI market?

As of September 2026, North America and Europe account for the entire disclosed agentic AI funding market in this dataset over the requested 12 months. North America produced 33 of 39 financings and raised $2.658B, while Europe produced 6 deals and raised $1.018B.

North America holds 84.62% of deal count and 72.30% of capital. Its average financing is $80.53M and its median is $38M, reflecting a much broader formation pipeline.

Europe holds only 15.38% of deals but 27.70% of dollars. Its average financing reaches $169.67M and its median $125M, producing a capital-share-to-deal-share ratio of 1.80x.

That apparent European financing advantage should not be generalized. Legora and Wonderful account for most of Europe's large rounds, so the regional average is driven by a small number of scale winners rather than a deeper market.

For a deeper geographic and company-level view, explore our Agentic AI market landscape.

Is the agentic AI opportunity set broad geographically or concentrated in one hub?

As of September 2026, the agentic AI opportunity set is geographically concentrated rather than broad over the requested 12 months. North America and Europe together account for 100% of qualifying disclosed capital and every financing in the dataset.

North America is the true breadth center because it produced 33 financings across application, infrastructure, memory and development-platform categories. Europe has fewer funded companies but several unusually large application winners.

No qualifying disclosed financings were found in Asia-Pacific, Latin America, the Middle East or Africa under the strict pure-player and disclosed-equity criteria. That absence is notable given the global distribution of AI engineering talent.

The dataset therefore measures venture-funded company formation rather than technical activity in general. A region can have strong AI research or implementation talent without producing a qualifying independent agentic AI financing during the observation window.

Is agentic AI a market of small experiments or scaled financings?

As of September 2026, the agentic AI market already contains a large number of scaled financings over the requested 12 months. Nineteen of 39 disclosed rounds were $50M or larger, while the median financing reached $43M.

Only 3 financings were below $5M, 6 were between $5M and below $20M, and 11 were between $20M and below $50M. The remaining 19 deals reached at least $50M.

Using the stricter megaround definition of more than $50M, 16 financings qualify, representing 41.03% of all deals. Twelve rounds exceeded $100M, equal to 30.77% of the complete disclosed sample.

The financing scale differs sharply by category. Vertical AI Agents have an $87.5M median, while Agent Memory Systems have a median of only $3.9M and Human Approval Agents remain represented by a single $4M round.

If you want more context on financing benchmarks and category maturity, see our Agentic AI funding report.

Who are the investors that appear the most in agentic AI fundraising?

As of September 2026, a relatively small group of venture firms repeatedly appears around the largest agentic AI companies over the requested 12 months. The clearest repeat names include Y Combinator, Bessemer Venture Partners, Sequoia Capital, Index Ventures, Insight Partners, General Catalyst, Benchmark, Redpoint, ICONIQ and Andreessen Horowitz.

Y Combinator appears across companies including Motion, Mem0, Giga, Legora, Concourse and AgentMail. Bessemer appears around Continua, Wonderful and multiple Legora financings, while Sequoia participates in Factory, Harvey and Parallel.

Index Ventures appears in Wonderful, Decagon and Parallel. General Catalyst, Benchmark, Redpoint and ICONIQ also recur, particularly around companies that moved through funding stages quickly.

Investor overlap is itself informative. Several of the largest agentic AI companies share syndicate members, so the number of financings should not be mistaken for an equal number of independent institutional views on the market.

Individual investor check sizes are generally not disclosed. These repeat counts therefore measure participation in qualifying rounds, not the exact amount of capital each investor personally committed.

Who is the author of this content?

NEW MARKET PITCH TEAM

We track new markets so founders and investors can move faster

We build living "market pitch" documents for emerging markets: AI, synthetic biology, new proteins, and more. Instead of outdated PDFs or hallucinated LLM answers, our clients get a clean, visual, always-updated view of what's really happening: key players, deals, regulations, and signals that matter. Learn more about us.