AI in Business

The Biggest US AI Startups Raising $100 Million or More in 2026

A smartphone displaying the word Venture, held in an office setting

This is a factual snapshot of the US-based AI startups that have raised $100 million or more in capital, or reached $100 million or more in annualized revenue, during 2026. The year has been defined by massive capital deployment into the sector, with OpenAI and Anthropic together accounting for $217 billion, or 43% of every venture dollar deployed worldwide, in the first half of the year. Furthermore, 88% of that AI capital went to US-based companies, underscoring the continued dominance of American innovation in this space.

Valuation and funding figures in the private market are fluid. “ARR” and “run-rate revenue” are used loosely and inconsistently by the industry. Some companies report annualized run-rate based on a single strong quarter, while others use trailing twelve-month figures. Additionally, valuations are private-market marks that can be stale by the time they are published, and figures are as of September 2026 and will move. This list captures the capital events and revenue milestones that defined the landscape through this period.

The companies below are organized by category to show where the money is actually going. From frontier labs with a combined valuation approaching $1.8 trillion (OpenAI’s $852 billion plus Anthropic’s $965 billion) to the physical AI startups building robots, these are the entities that have captured the most venture capital and attention in 2026.

Foundation models and frontier labs

The foundation model sector remains the most capital-intensive part of the AI industry. Two companies, OpenAI and Anthropic, have absorbed a staggering share of global venture capital, setting a new benchmark for what it means to be a “frontier” lab.

OpenAI

OpenAI remains the market leader in terms of capital raised. The company completed a massive $122 billion funding round in Q1 2026, achieving a post-money valuation of $852 billion. The round highlighted the immense confidence investors have in its large language model technologies, particularly through its ChatGPT product.

Anthropic

Anthropic has experienced one of the most rapid valuation escalations on record. In February 2026, the company raised $30 billion in a Series G round led by GIC and Coatue, valuing it at $380 billion. By May 2026, just three months later, it raised another $65 billion in a Series H led by Altimeter Capital, Dragoneer, Greenoaks, and Sequoia Capital. This second round pushed its post-money valuation to $965 billion—a roughly 2.5x increase in three months, which analysts noted as one of the steepest private valuation re-ratings on record. By the time of the Series H, Anthropic’s annualized run-rate revenue had crossed $47 billion. The company is best known for its Claude family of AI models.

xAI

xAI, the company founded by Elon Musk, raised $20 billion in a Series E round during 2026. The company is known for its Grok model, which is integrated into the X (formerly Twitter) platform. Unlike some of its peers, no specific post-money valuation figure was confirmed for this round in available sources, though the capital raise itself marks it as one of the largest funding events of the year.

Safe Superintelligence (SSI)

Founded in June 2024 by former OpenAI chief scientist Ilya Sutskever, SSI has moved quickly despite being a relatively young company. It has raised roughly $8 billion in total across multiple rounds. This included a $2 billion round in 2025 led by Greenoaks, with participation from Andreessen Horowitz, Lightspeed Venture Partners, and DST Global, followed by a $5 billion investment from Nvidia in July 2026 at a $32 billion valuation. As of August 2026, the roughly 50-person company had not shipped a commercial product, publicly released a model, or published research, which is highly unusual for a company at this valuation. A first release was rumored for August 2026, though no product launch had been independently confirmed at the time of writing.

AI infrastructure, chips and inference

As the demand for machine learning compute grows, infrastructure companies that build the chips and software to run these models have attracted significant capital. This category includes both chip manufacturers and software platforms that help enterprises deploy AI at scale.

Databricks

Databricks raised $5 billion in August 2026 at a $190 billion valuation. The company reported a $7 billion annualized revenue run-rate in Q2 2026, representing growth of more than 80% year over year. CEO Ali Ghodsi has indicated that an IPO is more likely in 2027 or later, suggesting the company is still in a growth phase rather than looking for an immediate public exit.

Cerebras Systems

Cerebras Systems, a maker of AI chips that compete directly with Nvidia, raised $1 billion in a Series H in February 2026 at a $23 billion post-money valuation. The company then went public on May 14, 2026, raising $5.6 billion in its IPO at an initial market cap of nearly $40 billion. This represented a 73% increase over its Series H valuation just three months earlier, signaling strong investor appetite for specialized AI hardware.

Groq

Groq, which makes AI inference chips, raised $650 million in June 2026 to expand its AI inference cloud data centers. This followed a deal in which Nvidia licensed Groq’s technology and hired away much of its talent. In August 2026, Groq raised a further $350 million round led by Disruptive, with planned Nvidia participation, at a $3.5 billion valuation. This valuation was roughly half what the company was worth about a year earlier, before the Nvidia deal, highlighting the volatility in the inference chip market.

Fireworks AI

Fireworks AI raised $1.505 billion in a Series D. The company provides enterprise AI inference tools, helping businesses deploy and run large language models efficiently. Its funding round underscores the continued demand for accessible, high-performance inference infrastructure.

Together AI

Together AI raised $800 million in a Series C. The company provides open-source AI infrastructure, catering to developers and enterprises that prefer open-weight models and flexible deployment options.

Baseten

Baseten raised $300 million in a Series E at a $5 billion valuation. The company provides AI inference infrastructure, focusing on making it easier for developers to serve and manage AI models in production environments.

Coding and developer tools

The coding tools sector has seen explosive growth, with AI-assisted development becoming a standard part of the software engineering workflow. Two companies in this category have reached extraordinary revenue and valuation milestones.

Cursor (built by Anysphere)

Cursor, an AI-assisted code editor built by Anysphere, reached roughly $4 billion in annualized revenue by May 2026, up from about $100 million just 16 months earlier. The company raised a $2.3 billion Series D in November 2025, co-led by Accel and Coatue, at a $29.3 billion valuation. In June 2026, SpaceX announced it would acquire Anysphere/Cursor in an all-stock deal valuing the company at approximately $60 billion. The deal closed on August 14, 2026. This means Cursor is technically no longer an independent startup, but its funding and revenue trajectory through the deal make it one of the defining AI stories of 2026.

Cognition

Cognition, the maker of the AI coding agent Devin, raised more than $1 billion in new funding in May 2026 at a $26 billion valuation. Devin represents the shift from code completion to autonomous coding agents, which can execute multi-step programming tasks independently.

Enterprise AI agents and platforms

Enterprises are moving beyond simple chatbots to deploy AI agents that can perform complex tasks across their internal systems. This category includes companies building platforms for customer service, enterprise search, and conversational AI.

Sierra

Sierra, co-founded by Bret Taylor, builds customer-service AI agents for enterprises. The company reached its first $100 million in annualized revenue in seven quarters, then added a second $100 million in just two more quarters, demonstrating rapid commercial adoption. Sierra raised a $950 million Series E in May 2026, led by GV and Tiger Global, at a $15.8 billion post-money valuation.

Glean

Glean is an enterprise AI search and assistant platform that helps employees find information across corporate data silos. The company was valued at $7.2 billion from a June 2025 Series F. By May 2026, Glean had crossed $300 million in annualized recurring revenue, reflecting the high demand for enterprise-grade AI search solutions.

Decagon

Decagon provides a conversational AI platform for enterprises. The company raised a $250 million Series D at a $4.5 billion valuation, indicating steady investor confidence in the enterprise conversational AI space.

Vertical AI companies focus on specific industries, tailoring their models and workflows to the unique needs of legal and healthcare professionals.

Harvey

Harvey is a legal AI platform that has raised $200 million in March 2026 at an $11 billion valuation, up from an $8 billion valuation in December 2025. The company reported around $190 million in annualized recurring revenue at the time, showing strong traction in the legal sector.

Assort Health

Assort Health is a healthcare AI and clinical-automation platform. The company raised a $120 million Series C at a $1.2 billion valuation, targeting the inefficiencies in clinical workflows.

Abridge

Abridge builds an AI medical-documentation platform used by clinicians to transcribe and summarize patient visits. The company has focused on reducing the administrative burden on doctors, though specific valuation figures vary and were not confirmed for this piece.

Robotics and physical AI

The “physical AI” category has seen massive capital inflows as investors bet on the commercialization of humanoid robots and foundation models for robotics.

Figure AI

Figure AI, a humanoid-robotics company, raised more than $1 billion in a Series C in September 2025 at a $39 billion post-money valuation. This was up roughly 15x from a $2.6 billion valuation in early 2024. Investors in the round included Nvidia, Microsoft, Intel Capital, Brookfield Asset Management, the OpenAI Startup Fund, and Jeff Bezos personally.

Physical Intelligence

Physical Intelligence builds foundation models for robots. The company raised roughly $1.1 billion in total across its first two years, including a $600 million Series B in November 2025 that valued the company at $5.6 billion.

Skild AI

Skild AI builds AI models specifically for robotics. The company raised $1.4 billion in a Series C at a $14 billion valuation, signaling strong interest in the software layer that will drive physical robots.

World Labs

World Labs, founded by Fei-Fei Li, builds “spatial intelligence” AI, which includes 3D and spatial world models. The company closed a $1 billion round on February 18, 2026, at a $5 billion valuation, up from a $1 billion valuation in September 2024. Total funding is close to $1.23 billion since its September 2024 seed round, backed by Autodesk, Nvidia, AMD, and Andreessen Horowitz. The company has not disclosed revenue.

Voice, video and creative AI

AI is transforming the creative industry, particularly in voice synthesis, speech recognition, and video understanding.

ElevenLabs

ElevenLabs is a voice AI platform that has seen explosive revenue growth. Its annualized revenue climbed from $100 million in January 2025 to $500 million by June 2025, to $1 billion by November 2025, and reached $2 billion by February 2026. This rapid growth highlights the commercial viability of high-fidelity voice AI.

Deepgram

Deepgram is a voice AI platform focused on speech recognition and transcription. The company raised a $130 million Series C at a $1.3 billion valuation, catering to enterprises that need accurate and scalable speech-to-text capabilities.

Search, data and consumer AI

The search and data landscape is being reshaped by AI-native companies that offer more than just keyword matching.

Perplexity

Perplexity is an AI search and “agent” company that closed a Series E-6 round at approximately $23 billion valuation on January 9, 2026. Its annualized revenue reached $750 million by August 2026, up from $232 million at the end of 2025. This growth was driven partly by the February 2026 launch of its “Comet”/agent product line and usage-based pricing. The company serves more than 100 million monthly active users, making it one of the largest consumer AI products by user base.

Twelve Labs

Twelve Labs builds AI systems for understanding and searching video content. The company raised a $100 million Series B, targeting the growing demand for video intelligence in media and enterprise applications.

Scale AI

Scale AI is a data-labeling and AI infrastructure company. In 2025, Meta invested approximately $14.3 billion for a 49% stake, valuing the company at $29 billion. Founder Alexandr Wang left to become Meta’s Chief AI Officer. Some major clients, reportedly including OpenAI, Google, and xAI, pulled back after the Meta deal, and the company’s 2026 revenue guidance fell to around $1 billion. New leadership, including interim CEO Jason Droege, has shifted the business toward helping large enterprises and public-sector clients build their own internal AI applications. The company emphasizes that it remains operationally independent from Meta despite the stake.

FAQ

How many AI startups are worth over $1 billion? While the exact number of AI startups worth over $1 billion fluctuates as private valuations are updated, the companies listed above represent a significant portion of the sector’s “unicorn” and “decacorn” class. Many of these companies have achieved valuations well above $10 billion, with OpenAI and Anthropic surpassing $800 billion.

What does ARR mean for an AI startup? ARR usually stands for Annual Recurring Revenue, a SaaS metric for predictable subscription income. In AI startup coverage, though, the term is often used loosely for “run-rate revenue” instead: current usage-based revenue for one month or quarter, multiplied out to a yearly figure. The two are not the same thing, and outlets rarely specify which one a headline number refers to, so treat any AI startup’s “ARR” as an estimate of current pace rather than a guarantee of next year’s revenue.

Are these AI startups independent companies? Most of the companies listed are independent, though some have significant corporate backing or strategic investments. For example, Scale AI has a 49% stake owned by Meta, and Cursor was acquired by SpaceX in August 2026, meaning it is no longer an independent startup.

What is the difference between foundation models and vertical AI? Foundation models, like those from OpenAI and Anthropic, are large-scale AI models trained on vast datasets that can be adapted for many different tasks. Vertical AI, like Harvey or Abridge, focuses on specific industries, such as law or healthcare, and is often fine-tuned or built specifically for the workflows of that sector.