Generative AI Startup Funding 2025-2026

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

The table below lists the 67 disclosed equity rounds in the generative AI dataset between August 2025 and September 2026. We define the generative AI market as revenue from products and services whose primary purpose is to create or transform text, code, images, audio, or video using generative models, including foundation-model APIs, GenAI tooling, GenAI-first applications, and related implementation services.

Each row shows the company, what it does, its category, the deal month, funding stage, round size, region, and main investors. For a broader view of companies, financing patterns, risks, and opportunities across this industry, see our Generative AI market report.

How active has fundraising been in the generative AI market?

As of September 2026, fundraising in the generative AI market has been exceptionally active over the past 12 months, with 67 disclosed equity rounds across 57 unique companies and approximately $203.79B of capital raised.

Deal flow averaged 4.79 disclosed rounds per month, with a median of 4.5. September 2025 was the busiest month in the dataset with 11 deals, while the partial September 2026 period contained one disclosed financing through September 2.

Dollar activity is much more volatile than deal activity. Average capital raised reached $14.56B per month, but the median monthly total was only $1.30B because OpenAI and Anthropic created enormous spikes in February and May 2026.

Removing rounds above $50M reduces disclosed capital from $203.79B to only $615.25M. That means aggregate funding is a poor proxy for the financing conditions faced by ordinary generative AI startups. For a broader view of current company formation and financing activity, see our Generative AI market analysis.

How concentrated has fundraising been in the generative AI market?

As of September 2026, fundraising in the generative AI market has been extraordinarily concentrated over the past 12 months. The largest financing alone represents 53.98% of all disclosed capital, while the top 3 rounds account for 92.25%.

Concentration remains extreme beyond the first few deals. The top 5 financings capture 94.36% of capital and the top 10 capture 96.63%, leaving only a small fraction of dollars for the remaining 57 transactions.

OpenAI's $110B financing is the single largest contributor. Anthropic's $65B and $13B rounds add another enormous layer of concentration, so headline market totals primarily describe the financing calendars of a few frontier-model companies.

This makes aggregate generative AI funding totals easy to misread. Capital growth can surge even when the number and size of ordinary startup rounds change very little, so concentration should always be examined alongside total dollars.

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

As of September 2026, most of the generative AI funding signal over the past 12 months is driven by outliers. 38 of 67 disclosed rounds exceeded $50M, and 33 exceeded $100M.

The mean disclosed round size is approximately $3.04B, while the median is $100M. That difference immediately shows how poorly the average describes a typical generative AI financing.

Removing rounds above $50M eliminates about 99.7% of disclosed capital, leaving only $615.25M. OpenAI alone contributes roughly 54% of the entire dataset, so even comparisons between months can become comparisons between individual companies.

The practical reading rule is simple: when generative AI funding rises sharply, first identify whether one frontier-model financing caused the move. Without that adjustment, market-wide momentum can be overstated.

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

As of September 2026, the generative AI market is broad on company formation but narrow on where the largest pools of capital go over the past 12 months. The dataset contains 57 unique companies across 67 disclosed financings.

Enterprise GenAI Applications produced 23 deals and GenAI Developer Platforms produced 21, together representing 65.7% of all transactions. Those companies span legal, medicine, customer service, IT, procurement, sales, coding, context, inference, and agent infrastructure.

The capital picture is radically narrower. Foundation Model APIs produced only 11 deals but captured 95.09% of disclosed dollars, meaning startup experimentation is broad while financing power is concentrated upstream.

Repeat fundraising reinforces this distinction. Companies such as Replit, Lovable, Legora, Harvey, Wonderful, OpenEvidence, Suno, and Higgsfield raised again within relatively short intervals, creating a visible group of companies separating from the wider formation pool.

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

As of September 2026, the generative AI market is overwhelmingly a late-stage scaling market when measured by capital over the past 12 months. Seed, Series A, and Series B together captured $5.89B, or only 2.89% of classified early- and late-stage capital.

Later stages captured approximately $197.74B, or 97.11%. Growth Equity alone represents $111.11B and Series D+ another $82.97B, overwhelmingly because the largest model companies raise at unprecedented scale.

Deal count tells a more balanced story. Seed produced 18 deals and Series A produced 15, meaning nearly half of all financings still occur close to company formation even though those rounds barely register in total capital.

Stage labels also require caution. Seed rounds range from ordinary single-digit millions to $480M for Humans& and $1.1B for Ineffable Intelligence, so stage name no longer reliably indicates capital intensity in frontier AI. We explore these financing dynamics further in our report on the Generative AI market.

Which categories attract the most investor attention in generative AI?

As of September 2026, Enterprise GenAI Applications attract the most investor attention by deal count in the generative AI market over the past 12 months, with 23 disclosed financings representing 34.33% of all deals.

GenAI Developer Platforms are close behind with 21 deals, or 31.34%. Together these two categories account for nearly two-thirds of all financing activity, showing that investors are exploring many application and infrastructure layers around the models themselves.

Foundation Model APIs account for only 11 deals, while Creative GenAI Tools account for 10. Governance is much smaller at 2 deals, and no qualifying pure-play GenAI Services financing appears in the verified dataset.

The application activity is fragmented across workflows rather than converging on one universal agent. Legal, customer support, medicine, IT, procurement, sales, commerce, and productivity each attract separate companies and capital pools. For more detail on the companies investors are backing across these layers, see our Generative AI market report covering funding and competition.

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

As of September 2026, Foundation Model APIs attract by far the most disproportionately large checks in the generative AI market over the past 12 months. The category captures 95.09% of capital from only 16.42% of disclosed deals.

That creates a capital-share-to-deal-share ratio of 5.792x. No other generative AI category comes remotely close to parity between its share of dollars and its share of transactions.

Creative GenAI Tools have a ratio of only 0.082x, GenAI Developer Platforms 0.073x, Enterprise GenAI Applications 0.040x, and GenAI Governance Tools 0.003x. Those figures show just how completely frontier-model rounds dominate capital allocation.

Average round size tells the same story. Foundation Model APIs average approximately $17.62B per financing with a $1.10B median, compared with $222.43M for developer platforms, $122.92M for enterprise applications, and $249.10M for creative tools.

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

As of September 2026, North America matters most for fundraising in the generative AI market over the past 12 months. It accounts for 46 of 67 disclosed deals and approximately $197.98B of capital.

That represents 68.66% of transactions but 97.15% of dollars. The gap is mostly explained by enormous financings for OpenAI, Anthropic, and other US-based model and platform companies rather than an absence of company formation elsewhere.

Europe generated 14 deals, or 20.90% of activity, and raised approximately $5.46B. The Middle East contributed 4 deals and $290M, while Asia-Pacific contributed 3 deals and $57M.

Latin America and Africa contributed no qualifying disclosed financings in this strict dataset. Regional comparisons therefore need both deal share and capital share, because frontier-model rounds radically distort the dollar picture. For a deeper regional view, see our Generative AI market report by geography.

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

As of September 2026, the generative AI opportunity set is geographically broad enough to produce companies across several regions, but financing remains strongly concentrated in North America over the past 12 months.

North America produces 68.66% of deals, so nearly one-third of disclosed company financings still occur elsewhere. Europe alone represents 20.90% of transactions, demonstrating meaningful formation outside the US ecosystem.

The dollar distribution is much narrower. North America captures 97.15% of disclosed capital, while Europe receives 2.68%, the Middle East 0.14%, and Asia-Pacific just 0.03%.

Europe's median disclosed round is actually $175M versus $130M in North America. Its capital deficit therefore comes primarily from lacking $10B to $100B frontier-model financings, not from every European company raising smaller rounds.

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

As of September 2026, the generative AI market is dominated by scaled financings rather than small experiments over the past 12 months. 38 of 67 disclosed deals exceeded $50M, representing 56.72% of all transactions.

33 rounds exceeded $100M, meaning 49.25% of all disclosed financings were above that threshold. Exactly $50M and $100M rounds are excluded from those strict greater-than calculations.

At the smaller end, only 3 deals were $5M or less, 12 were above $5M through $20M, and 14 were above $20M through $50M. Large financings therefore outnumber every smaller size bucket individually.

The median round is $100M and the average is approximately $3.04B. The median is more useful, but even it reflects a strict pure-play universe containing many scaled model, coding, legal, voice, video, and agent companies. For more context on how these financing sizes affect competitive dynamics, see our full Generative AI market deck.

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

As of September 2026, the investors appearing most repeatedly in generative AI fundraising over the past 12 months include Lightspeed Venture Partners, Sequoia Capital, Andreessen Horowitz, Coatue, General Catalyst, Nvidia, Accel, GV, DST Global, Insight Partners, Menlo Ventures, and Bessemer Venture Partners.

Lightspeed appears across model companies, developer infrastructure, and applications including Anthropic, Mistral AI, Paid, Emergent, Inferact, Sycamore, and Sandstone. Sequoia similarly spans Anthropic, Humans&, Sesame, Factory, Harvey, and ElevenLabs.

Andreessen Horowitz appears across Mistral AI, Replit, Anysphere, Harvey, Inferact, Lio, and Probably, while Coatue appears across Anthropic, Replit, Anysphere, Harvey, and Sycamore. The pattern suggests leading investors prefer portfolio exposure across multiple generative AI layers.

Nvidia is strategically distinctive because its recurring investments reach beyond semiconductor companies into model labs, developer platforms, and creative applications. That gives it equity exposure to businesses generating demand for the compute infrastructure underlying generative AI.

One caveat matters for every investor ranking: announcements usually disclose selected participants rather than exact check sizes or complete investor lists. Repeat-investor counts should therefore be interpreted as conservative observed participation, not a complete league table of capital 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.