OpenAI and Anthropic claim 80% of Forbes AI 50 funding as a new wave of pre-unicorn challengers emerges

Two annual Forbes rankings published this year put hard numbers on what enterprise buyers have been sensing for months: AI is no longer a bet on future potential. It is a measurable, revenue-generating reality concentrated at the top and still proliferating aggressively at the bottom.

Forbes' eighth annual AI 50 list, edited by Rashi Shrivastava and published in April, tallied $305.6 billion in total venture funding across its 50 honorees. OpenAI and Anthropic together account for $242.6 billion of that sum, according to Forbes, roughly 80% of the entire pool. At the end of February, OpenAI reportedly posted more than $25 billion in annualized revenue. Anthropic, for its part, said its revenue run rate crossed $30 billion in early April.

What the concentration of capital means for enterprise buyers

For CIOs and procurement directors evaluating foundation model vendors, those revenue figures matter more than valuations. A platform generating $25 to $30 billion in annualized run-rate revenue has the financial durability to sustain enterprise contracts, maintain roadmaps, and absorb the support costs that come with large-scale deployments. That bar is now set in public.

Below the two giants, the AI 50 reveals a second tier of enterprise-relevant platforms with specific functional footprints. French startup Mistral, according to Forbes, is selling its open-weight models to large corporations including Cisco and to European government agencies, a positioning that speaks directly to procurement teams with data-residency or sovereignty requirements. Physical Intelligence, a San Francisco-based company, has raised $1 billion to train foundational models for robotics, collecting data from human teleoperators in environments like kitchens and manufacturing-adjacent spaces. For operations leaders watching the autonomous-robotics market, it is one of the more concrete early bets.

The coding category is evolving with particular speed. Anthropic's Claude Code and OpenAI's Codex are both pushing into developer tooling, according to Forbes, putting pressure on coding-focused startups like Cursor, valued at $29.3 billion. Cognition, a $10 billion-valued coding agent startup that debuted on this year's AI 50, acquired the remaining assets of Windsurf after Google had already paid $2.4 billion to hire Windsurf's cofounders and license its technology. Enterprise software and DevOps teams evaluating AI coding tools are now navigating a market reshaped by both acquisition and direct competition from the labs themselves.

Vertical-specific AI companies are closing in on unicorn status

The second Forbes list, the Next Billion-Dollar Startups 2026, published July 28 and co-produced with TrueBridge Capital Partners, adds the forward-looking dimension. The list has a credible track record: of 275 alumni over its 12-year run, 60% became unicorns, according to Forbes, with graduates including Duolingo and DoorDash. Nearly half of last year's picks already exceeded a $1 billion valuation.

This year, Forbes notes that nearly all companies on the list use AI in some fashion, spanning industries from bone marrow research to biological threat detection and cybersecurity. That breadth matters for enterprise sourcing teams because it signals that AI-native alternatives are arriving in virtually every vendor category, not just software.

Two companies on the Next Billion-Dollar list illustrate the operational range. American Terawatt, a San Francisco startup that has raised $52 million at a $350 million valuation, is building direct-current power grids specifically for AI data centers, arguing that eliminating the AC-to-DC conversion step recovers meaningful energy that is currently lost. For data center operators and infrastructure procurement teams, that proposition connects a software-era problem (AI compute density) to a hardware-era solution (grid redesign). Abby Care, backed by Sequoia Capital, Thrive Capital, and Khosla Ventures, has raised $45 million at a $225 million valuation and is using AI to help families of disabled or elderly people become paid Medicaid caregivers, with an app that handles timesheets, charting, and an AI-powered assistant. Healthcare system operators and workforce management buyers will want to track how the Medicaid-funded caregiver marketplace model scales.

Market consolidation and the newcomers pushing back

The AI 50 also documented a year of notable consolidation. Three companies from the 2025 list were absorbed or restructured by larger players: Elon Musk's xAI was acquired by SpaceX, creating a combined entity Forbes valued at $1.25 trillion; Google paid $2.4 billion to hire Windsurf's cofounders and license its technology; and Scale AI's CEO Alexandr Wang departed to lead Meta's superintelligence lab, though Scale AI says it remains independent and has seen revenue growth since.

Against that consolidation, 20 newcomers joined the AI 50 this year. Gamma, an AI presentation builder valued at $2.1 billion, crossed $100 million in annualized revenue with just 50 employees, according to Forbes. Chai Discovery, a two-year-old company valued at $1.3 billion, is applying AI to drug development. Rogo, based in New York, reports that roughly 25,000 bankers and investors use its software. Each of those figures gives procurement and vendor evaluation teams a concrete point of reference when assessing whether an AI startup has genuine commercial traction.

Taken together, the two Forbes lists draw a clear operational picture: the foundation model market is consolidating around a small number of highly capitalized platforms, while a broad and growing class of AI-native, vertical-specific companies is approaching unicorn thresholds across healthcare, infrastructure, financial services, and developer tooling. Enterprise teams that have been waiting for the AI vendor market to stabilize before committing will find that the top tier already has; the question now is which second-tier platforms mature fast enough to earn long-term contracts.