Compiled: January 2025
Total Companies: 100 Genuine AI-Native Startups
Combined Valuation: $1.62 Trillion
Combined Funding: $234.2 Billion
This report presents the top 100 funded AI-native startups, ranked by valuation. These companies represent the genuine infrastructure, applications, and foundation models powering the AI revolution - not pretenders using AI as a marketing buzzword.
| Metric | Value |
|---|---|
| Total Companies | 100 |
| Combined Valuation | $1,616.5 Billion |
| Combined Funding | $234.2 Billion |
| Average Valuation | $16.2 Billion |
| Average Funding | $2.3 Billion |
| US Companies | 84 (84%) |
| International | 16 (16%) |
✅ INCLUDED:
- Foundation models (OpenAI, Anthropic, Mistral)
- AI infrastructure (Together AI, CoreWeave, Lambda)
- AI-native applications (Cursor, Perplexity, Harvey)
- AI developer tools (LangChain, LlamaIndex)
- AI robotics (Figure, Physical Intelligence)
- Generative AI (Midjourney, ElevenLabs, Runway)
❌ EXCLUDED:
- C3.ai, IBM Watson, Salesforce Einstein (AI pretenders)
- Companies using AI as just a feature
- Consultancies claiming AI expertise
- Traditional software with AI marketing
Valuation: $500.0B | Funding: $59.0B | Location: San Francisco, CA, USA
Team Size: ~750 employees | Category: Foundation Model
Product: ChatGPT, GPT-4/5, DALL-E, Sora, Codex - Leading generative AI platform powering conversational AI, image/video generation, and enterprise APIs
Key Competitors: Anthropic (Claude), Google DeepMind (Gemini), Meta AI (Llama), Mistral AI
Bull Case: First-mover advantage with 500M+ weekly active users; $12.7B projected 2025 revenue; Microsoft's strategic backing; dominant mindshare in consumer and enterprise AI
What Paul Graham Would Advise: You've won the first battle but the war is just beginning. Don't become the IBM of AI - stay scrappy. Focus on what doesn't scale: personally onboard your biggest enterprise customers. And remember, growth is the only metric that matters - but make sure it's sustainable growth, not just hype.
Notable Investors: SoftBank, Thrive Capital, Microsoft, Tiger Global, Sequoia Capital
Valuation: $230.0B | Funding: $38.0B | Location: San Francisco, CA, USA
Team Size: ~200 employees | Category: Foundation Model
Product: Grok chatbot and AI models integrated with X platform, featuring real-time access to social media data and unfiltered responses
Key Competitors: OpenAI (ChatGPT), Anthropic (Claude), Google (Gemini), Perplexity
Bull Case: Elon Musk's star power; unique real-time data access via X; $20B+ in compute resources; potential to disrupt with 'anti-woke' positioning
What Paul Graham Would Advise: You're playing catch-up, which means you need to be relentlessly resourceful. Find the schlep that OpenAI won't do - maybe that's truly open models or radical transparency. Live in the future you want to create, not the past you're trying to beat.
Notable Investors: Valor Equity Partners, Fidelity, Qatar Investment Authority, Sequoia Capital, Andreessen Horowitz
Valuation: $183.0B | Funding: $26.0B | Location: San Francisco, CA, USA
Team Size: ~2,847 employees | Category: Foundation Model
Product: Claude AI assistant and Claude Code - Enterprise-focused LLM with emphasis on AI safety, Constitutional AI, and long context windows
Key Competitors: OpenAI, Google DeepMind, Cohere, AI21 Labs
Bull Case: AI safety-first approach resonates with enterprises; Constitutional AI methodology; $60B+ valuation growth shows market confidence; preferred by developers for coding tasks
What Paul Graham Would Advise: Your safety-first approach is your moat - lean into it. Do things that don't scale: personally help companies implement responsible AI. Remember, the best startup ideas come from solving your own problems - you built Claude because you cared about safe AI.
Notable Investors: Amazon, Google, Lightspeed Venture Partners, Spark Capital, Salesforce Ventures
Valuation: $134.0B | Funding: $21.0B | Location: San Francisco, CA, USA
Team Size: ~12,312 employees | Category: AI Infrastructure/Data
Product: Data Intelligence Platform - Unified analytics, ML, and generative AI platform built on Apache Spark with Lakehouse architecture
Key Competitors: Snowflake, Cloudera, AWS, Google Cloud, Microsoft Azure
Bull Case: Open-source foundation drives adoption; 60%+ YoY growth; $4.8B+ revenue run rate; 60%+ of Fortune 500 as customers; acquired MosaicML for $1.3B
What Paul Graham Would Advise: You're becoming the default data platform for AI - that's powerful. But don't rest on your laurels. Keep doing things that don't scale: help customers migrate personally. The companies that win are the ones that grow fastest, and growth comes from making customers insanely happy.
Notable Investors: Andreessen Horowitz, Coatue, Tiger Global, T. Rowe Price, Microsoft
Valuation: $45.0B | Funding: $11.0B | Location: Mountain View, CA, USA
Team Size: ~2,500 employees | Category: Autonomous Vehicles
Product: Self-driving car technology and robotaxi service operating in select cities with fully autonomous vehicles
Key Competitors: Tesla (FSD), Cruise (GM), Aurora, Zoox (Amazon), Pony.ai
Bull Case: Only company with commercial robotaxi service; 2M+ autonomous miles driven; Alphabet backing provides unlimited resources; first-mover in autonomous ride-hailing
What Paul Graham Would Advise: Hardware is hard - you've been at this for 15+ years. The schlep is the moat. Keep iterating, even when progress feels slow. Default alive means having the runway to survive until the technology is ready. You're living in the future - now make it profitable.
Notable Investors: Alphabet (Google), Andreessen Horowitz, AutoNation, Magna International
Valuation: $39.0B | Funding: $2.0B | Location: Sunnyvale, CA, USA
Team Size: ~200 employees | Category: AI Robotics
Product: Humanoid robots powered by AI for commercial and industrial applications - Figure 01 and Figure 02 models
Key Competitors: Tesla (Optimus), Boston Dynamics, Agility Robotics, 1X Technologies, Apptronik
Bull Case: Partnership with BMW for factory deployment; OpenAI collaboration for AI brain; $2B+ valuation with minimal funding shows capital efficiency; physical AI is the next frontier
What Paul Graham Would Advise: You're tackling one of the hardest problems in tech. The schlep blindness around robotics is your opportunity. Do things that don't scale: manually train each robot in early deployments. Remember, startups die when they lose momentum - keep shipping robots.
Notable Investors: Microsoft, OpenAI Startup Fund, NVIDIA, Jeff Bezos, Parkway Venture Capital
Valuation: $32.0B | Funding: $3.0B | Location: Palo Alto, CA, USA
Team Size: ~50 employees | Category: Foundation Model
Product: AI safety research and superintelligence development - founded by Ilya Sutskever after leaving OpenAI
Key Competitors: OpenAI, Anthropic, DeepMind, Thinking Machines Lab
Bull Case: Ilya Sutskever's pedigree (co-founder of OpenAI); $32B valuation with no product shows investor confidence; safety-first approach may become regulatory moat; attracting top AI talent
What Paul Graham Would Advise: You have the credibility to tackle the hardest problem in AI - safety. But remember, startups need to ship. Don't let perfect be the enemy of good. Build something people want first, then make it safe. The best way to ensure AI safety is to be the one building it.
Notable Investors: Andreessen Horowitz, Sequoia Capital, DST Global, SV Angel
Valuation: $31.0B | Funding: $6.0B | Location: Costa Mesa, CA, USA
Team Size: ~2,800 employees | Category: AI Defense
Product: AI-powered defense technology including autonomous drones, border security systems, and battlefield management software
Key Competitors: Palantir, Lockheed Martin, Raytheon, Northrop Grumman, Shield AI
Bull Case: Software-first approach disrupting legacy defense contractors; major contracts with US military and allies; Palmer Luckey's track record; autonomous systems are the future of warfare
What Paul Graham Would Advise: Defense tech is the ultimate schlep - that's your moat. Do things that don't scale: personally embed with military units to understand their problems. Live in the future where software eats defense. Remember, the best products come from solving real problems for real people.
Notable Investors: Founders Fund, Andreessen Horowitz, Valor Equity Partners, Elad Gil
Valuation: $29.3B | Funding: $2.3B | Location: San Francisco, CA, USA
Team Size: ~300 employees | Category: AI Coding
Product: AI-powered code editor (VS Code fork) with intelligent code completion, generation, and understanding - the 'vibe coding' platform
Key Competitors: GitHub Copilot, Amazon CodeWhisperer, Replit, JetBrains AI
Bull Case: $1B+ ARR in record time; fastest-growing B2B tool ever; Jensen Huang's 'favorite enterprise AI service'; used by OpenAI, Stripe, Spotify engineers; $29B valuation in 6 months
What Paul Graham Would Advise: You've found product-market fit - now don't screw it up. Do things that don't scale: personally help the biggest engineering teams adopt Cursor. Remember, growth is the only metric that matters. Keep iterating based on what your users tell you, not what you think they need.
Notable Investors: Accel, Coatue, Thrive Capital, NVIDIA, Google Ventures
Valuation: $29.0B | Funding: $14.3B | Location: San Francisco, CA, USA
Team Size: ~1,200 employees | Category: AI Training Data
Product: Data labeling, annotation, and RLHF platform for training AI models - the 'picks and shovels' of the AI gold rush
Key Competitors: Labelbox, Snorkel AI, Sama, Appen, Amazon SageMaker Ground Truth
Bull Case: Meta acquired 49% stake for $14.3B; $2B+ revenue; 'backbone of intelligent systems'; powers OpenAI, Google, Meta training; essential infrastructure for all AI development
What Paul Graham Would Advise: You're the infrastructure play everyone needs. But don't get complacent. Do things that don't scale: manually label data for your first customers to understand the problem deeply. Remember, the best businesses solve painful problems that others avoid.
Notable Investors: Meta (49% stake), Accel, Index Ventures, Founders Fund, Tiger Global
Valuation: $20.0B | Funding: $1.5B | Location: San Francisco, CA, USA
Team Size: ~250 employees | Category: AI Search
Product: AI-powered search engine with conversational answers, real-time information, and cited sources - challenging Google
Key Competitors: Google Search, Bing AI, OpenAI SearchGPT, You.com, Brave Search
Bull Case: 780M+ queries/month; $150M+ ARR; real-time AI search with citations; challenging Google's monopoly; potential Chrome acquisition bid; preferred by researchers
What Paul Graham Would Advise: You're David vs Goliath - that's exactly where startups win. Do things that don't scale: personally reach out to every power user. Remember, the best startup ideas come from solving your own problems. You built Perplexity because you hated search - now make everyone else love it.
Notable Investors: NVIDIA, Jeff Bezos, IVP, NEA, Elad Gil, Nat Friedman
Valuation: $19.0B | Funding: $4.0B | Location: Roseland, NJ, USA
Team Size: ~800 employees | Category: AI Infrastructure
Product: GPU cloud infrastructure provider specializing in NVIDIA GPUs for AI/ML workloads - the 'AWS for AI compute'
Key Competitors: Lambda Labs, Together AI, AWS, Google Cloud, Azure, Crusoe Cloud
Bull Case: Massive GPU supply agreements with NVIDIA; $1.9B revenue run rate; IPO-bound; Microsoft as major customer; solving the compute bottleneck for AI
What Paul Graham Would Advise: You're in a commodity business - differentiate through service. Do things that don't scale: personally help customers optimize their GPU workloads. Remember, the companies that win are the ones that make customers happiest. Be relentlessly resourceful in securing GPU supply.
Notable Investors: Magnetar Capital, NVIDIA, Fidelity, Coatue, Jane Street
Valuation: $14.0B | Funding: $3.0B | Location: Paris, France
Team Size: ~150 employees | Category: Foundation Model
Product: Le Chat AI assistant and open-source LLMs - Europe's leading AI lab championing 'AI sovereignty'
Key Competitors: OpenAI, Anthropic, Cohere, Google DeepMind, Meta AI
Bull Case: Europe's AI champion with government backing; open-source approach attracts developers; Macron endorsement; $14B valuation in 18 months; alternative to US AI dominance
What Paul Graham Would Advise: You're the European alternative - lean into that. Do things that don't scale: personally onboard European enterprises worried about data sovereignty. Remember, the best businesses solve real problems - European companies need AI that respects their privacy laws.
Notable Investors: Microsoft, Lightspeed Venture Partners, Andreessen Horowitz, General Catalyst, Bpifrance
Valuation: $12.0B | Funding: $2.0B | Location: San Francisco, CA, USA
Team Size: ~100 employees | Category: Foundation Model
Product: AI research lab focused on developing advanced AI systems - founded by former OpenAI researchers
Key Competitors: OpenAI, Anthropic, Safe Superintelligence, xAI
Bull Case: Founded by top OpenAI talent; $12B valuation with minimal funding; attracting researchers frustrated with OpenAI's commercialization; research-first approach
What Paul Graham Would Advise: You're building the next-generation research lab. Don't let fundraising success distract from the mission. Do things that don't scale: personally recruit every researcher. Remember, the best teams build the best products - focus on team quality above all else.
Notable Investors: Conviction (Sarah Guo), Andreessen Horowitz, NVIDIA, Sequoia Capital
Valuation: $11.0B | Funding: $0.3B | Location: San Francisco, CA, USA
Team Size: ~1,000 employees | Category: AI Productivity
Product: All-in-one workspace with AI-powered writing, summarization, and knowledge management - Notion AI integrated throughout
Key Competitors: Microsoft Loop, Coda, Obsidian, Roam Research, Asana
Bull Case: $300M+ ARR; 30M+ users; capital efficient with minimal funding; AI features driving premium upgrades; becoming the default knowledge OS for startups
What Paul Graham Would Advise: You've built something people love with almost no funding - that's the dream. Now add AI thoughtfully, not as a gimmick. Do things that don't scale: personally help power users build their workflows. Remember, the best products feel like extensions of your mind.
Notable Investors: Sequoia Capital, Index Ventures, Coatue, First Round Capital
Valuation: $10.0B | Funding: $0.6B | Location: Dallas, TX, USA
Team Size: ~300 employees | Category: AI Bio
Product: AI-powered de-extinction technology to resurrect woolly mammoths, dire wolves, and other extinct species
Key Competitors: None direct; competes with traditional conservation, synthetic biology companies
Bull Case: Unique mission capturing public imagination; George Church's involvement; $10B valuation for 'Jurassic Park' company; potential applications in conservation and healthcare
What Paul Graham Would Advise: You're tackling the ultimate schlep - bringing back extinct species. That's your moat. Do things that don't scale: personally work with every conservation partner. Remember, the best startups solve problems that seem impossible. Live in the future you want to create.
Notable Investors: Thomas Tull, Paris Hilton, Tony Robbins, Winklevoss Capital, Bold Capital
Valuation: $10.0B | Funding: $0.3B | Location: San Francisco, CA, USA
Team Size: ~200 employees | Category: AI Agents
Product: AI-powered customer service agents that can autonomously resolve complex customer issues
Key Competitors: Intercom, Zendesk AI, Freshworks, Moveworks, Ada
Bull Case: $635M raised; Bret Taylor (ex-Salesforce co-CEO) founding team; enterprise-grade AI agents; solving the $500B customer service market
What Paul Graham Would Advise: Customer service is the ultimate schlep - perfect for AI disruption. Do things that don't scale: manually train agents for your first customers. Remember, the best AI products don't replace humans, they augment them. Make your customers' customers happier.
Notable Investors: Sequoia Capital, Benchmark, Conviction, Greenoaks, Thrive Capital
Valuation: $10.0B | Funding: $0.3B | Location: San Francisco, CA, USA
Team Size: ~100 employees | Category: AI Recruiting
Product: AI-powered talent marketplace and recruiting platform using AI to match candidates with jobs
Key Competitors: LinkedIn, Indeed, Greenhouse, Lever, Ashby
Bull Case: $10B valuation from $250M valuation in 12 months; AI-native approach to recruiting; solving the $200B talent market; young founders (early 20s)
What Paul Graham Would Advise: Recruiting is broken - you're fixing it with AI. Do things that don't scale: personally interview your first 100 placements. Remember, the best businesses solve problems you personally experience. Keep iterating until both candidates and employers love you.
Notable Investors: Benchmark, Felicis Ventures, General Catalyst, Peter Thiel
Valuation: $9.8B | Funding: $0.9B | Location: San Francisco, CA, USA
Team Size: ~200 employees | Category: AI Coding
Product: Devin - AI software engineer that can autonomously write, debug, and deploy code; Windsurf IDE
Key Competitors: GitHub Copilot, Cursor, Amazon CodeWhisperer, Poolside, Magic
Bull Case: First 'AI software engineer'; viral demo (30M+ views); $10.2B valuation; acquired Windsurf for $1.25B; $73M ARR; autonomous coding agents are the future
What Paul Graham Would Advise: You've created the first true AI engineer - now make it work for real teams. Do things that don't scale: personally pair program with early adopters. Remember, the best AI products feel like magic. Keep shipping until every developer wants a Devin.
Notable Investors: Founders Fund, Conviction, 8VC, Elad Gil, Nat Friedman
Valuation: $9.0B | Funding: $0.7B | Location: New York, NY, USA
Team Size: ~500 employees | Category: AI Security
Product: AI-powered data security platform for discovering, protecting, and managing sensitive data across cloud environments
Key Competitors: Varonis, BigID, Securiti, OneTrust, Imperva
Bull Case: $9B valuation; $100M+ ARR; AI-native approach to data security; solving the $50B data protection market; Israeli founders with military cybersecurity background
What Paul Graham Would Advise: Data security is the ultimate schlep - that's your opportunity. Do things that don't scale: personally audit your first customers' data. Remember, the best security products make customers sleep better at night. Be the company they trust with their most sensitive data.
Notable Investors: Accel, Sequoia Capital, Coatue, Redpoint Ventures, Georgian
Valuation: $8.0B | Funding: $0.5B | Location: San Francisco, CA, USA
Team Size: ~350 employees | Category: AI Legal
Product: AI platform for legal professionals - document review, contract drafting, legal research, and due diligence
Key Competitors: Casetext (Thomson Reuters), Legora, Luminance, CoCounsel, LexisNexis
Bull Case: $100M+ ARR; 337+ law firm clients including top firms; lawyers guiding AI development; $8B valuation; OpenAI partnership; transforming the $800B legal industry
What Paul Graham Would Advise: Legal is the perfect vertical for AI - tons of documents, high stakes, expensive humans. Do things that don't scale: personally onboard the first 10 law firms. Remember, the best vertical AI is built by domain experts. Hire lawyers who can code.
Notable Investors: Sequoia Capital, Andreessen Horowitz, OpenAI Startup Fund, Conviction, Elad Gil
Valuation: $7.2B | Funding: $1.0B | Location: San Francisco, CA, USA
Team Size: ~1,500 employees | Category: AI Sales
Product: AI-powered revenue intelligence platform that analyzes sales calls, emails, and meetings to improve win rates
Key Competitors: Chorus (ZoomInfo), Clari, Outreach, Salesloft, Wingman
Bull Case: $200M+ ARR; 4,000+ customers; category creator for revenue intelligence; AI analyzes billions of sales interactions; transforming how B2B sales works
What Paul Graham Would Advise: You've created a new category - now own it. Do things that don't scale: personally review sales calls with your first customers. Remember, the best products solve real pain points. Every sales leader wants to know why they lose deals - you tell them.
Notable Investors: Sequoia Capital, Battery Ventures, Coatue, NextWorld Capital, Norwest Venture Partners
Valuation: $6.3B | Funding: $0.5B | Location: Redwood City, CA, USA
Team Size: ~400 employees | Category: AI Training
Product: Programmatic data labeling and ML platform using weak supervision to automate training data creation
Key Competitors: Scale AI, Labelbox, Appen, Amazon SageMaker Ground Truth
Bull Case: Stanford research background; weak supervision approach; $148M revenue; Fortune 500 customers; solving the data bottleneck for enterprise AI
What Paul Graham Would Advise: Data labeling is the schlep everyone avoids - that's your moat. Do things that don't scale: manually label data for your first customers to understand the pain. Remember, the best businesses solve annoying problems that others won't touch.
Notable Investors: Greylock Partners, Lightspeed Venture Partners, GV, In-Q-Tel, Addition
Valuation: $6.3B | Funding: $1.0B | Location: Boston, MA, USA
Team Size: ~1,000 employees | Category: AI ML Platform
Product: Enterprise AI platform for building, deploying, and managing machine learning models at scale
Key Competitors: H2O.ai, Databricks, Dataiku, AWS SageMaker, Google Vertex AI
Bull Case: Pioneer in AutoML; enterprise focus; hundreds of Fortune 1000 customers; making AI accessible to non-data scientists; $100M+ ARR
What Paul Graham Would Advise: Democratizing AI is a noble mission. Do things that don't scale: personally help your first enterprise customers build models. Remember, the best products make complex things simple. Keep removing friction until anyone can build AI.
Notable Investors: New Enterprise Associates, Sapphire Ventures, Meritech Capital Partners, DFJ Growth
Valuation: $6.0B | Funding: $2.0B | Location: Mountain View, CA, USA
Team Size: ~1,200 employees | Category: Autonomous Vehicles
Product: Autonomous delivery vehicles for last-mile logistics - compact self-driving cars for goods, not people
Key Competitors: Waymo, Amazon Scout, Starship Technologies, KiwiBot, Refraction AI
Bull Case: Ex-Google self-driving team; partnerships with Domino's, Walmart, Kroger; $6B valuation; focused on easier problem (goods vs people); regulatory path clearer than robotaxis
What Paul Graham Would Advise: Delivery is the right first market for autonomy - lower stakes than passengers. Do things that don't scale: personally map every delivery route. Remember, the best startups start with a narrow use case and expand. Own delivery, then expand.
Notable Investors: SoftBank Vision Fund, Greylock Partners, Gaorong Capital, Tiger Global, Fidelity
Valuation: $6.0B | Funding: $1.0B | Location: San Francisco, CA, USA
Team Size: ~150 employees | Category: AI Robotics
Product: AI models and learning algorithms for robot control - bringing general-purpose intelligence to physical robots
Key Competitors: Covariant, Sanctuary AI, Figure, Tesla Optimus, 1X Technologies
Bull Case: $5.6B valuation; $1B+ raised; Jeff Bezos and Thrive Capital backing; former Google Brain researchers; 'GPT moment for robotics' potential
What Paul Graham Would Advise: Robotics + AI is the next frontier. Do things that don't scale: personally train robots for your first customers. Remember, the best AI companies collect the best data. Every robot deployment is a data collection opportunity.
Notable Investors: CapitalG, Thrive Capital, Jeff Bezos, OpenAI, Lux Capital
Valuation: $6.0B | Funding: $0.8B | Location: Cambridge, MA, USA
Team Size: ~200 employees | Category: AI Healthcare
Product: AI-powered clinical decision support platform providing evidence-based answers for physicians at point of care
Key Competitors: UpToDate, DynaMed, IBM Watson Health, Infermedica
Bull Case: 40% of US physicians use the platform; $12B valuation; Sequoia and Thrive backing; solving the $4T healthcare information problem; FDA breakthrough device designation
What Paul Graham Would Advise: Healthcare is the ultimate schlep - but the impact is massive. Do things that don't scale: personally shadow physicians to understand their workflow. Remember, the best healthcare AI makes doctors better, not replaces them. Save lives, then make money.
Notable Investors: Sequoia Capital, Thrive Capital, GV, General Catalyst, Andreessen Horowitz
Valuation: $5.5B | Funding: $1.1B | Location: Toronto, Canada
Team Size: ~843 employees | Category: Foundation Model
Product: Enterprise LLM platform with Command family of models - cloud-agnostic deployment for businesses
Key Competitors: OpenAI, Anthropic, AI21 Labs, Google Gemini
Bull Case: $100M+ ARR; enterprise-focused; cloud-agnostic; Canadian government backing ($240M); strong NLP research team from Google Brain; privacy-first approach
What Paul Graham Would Advise: Enterprise AI is your moat - lean into it. Do things that don't scale: personally deploy models for your first enterprise customers. Remember, the best B2B products are built by understanding customer needs deeply. Be the AI partner enterprises trust.
Notable Investors: Index Ventures, Tiger Global, NVIDIA, Salesforce Ventures, Inovia Capital
Valuation: $5.3B | Funding: $1.0B | Location: San Diego, CA, USA
Team Size: ~800 employees | Category: AI Defense
Product: AI-powered autonomous systems for defense - drones and aircraft that operate without GPS or communications
Key Competitors: Anduril, Skydio, Palantir, General Atomics, Elbit Systems
Bull Case: $5.3B valuation; $240M Series F; Hivemind AI pilot; US military contracts; former Navy SEAL founders; solving critical national security problem
What Paul Graham Would Advise: Defense tech requires trust above all. Do things that don't scale: personally embed with military units. Remember, the best defense products save lives. Be the company warfighters trust when comms are down.
Notable Investors: Andreessen Horowitz, Point72 Ventures, Breyer Capital, Homebrew, Snowpoint Ventures
Valuation: $5.1B | Funding: $0.5B | Location: San Francisco, CA, USA
Team Size: ~600 employees | Category: AI Security
Product: AI-powered email security platform that detects and stops sophisticated email attacks
Key Competitors: Proofpoint, Mimecast, Barracuda, Tessian, Armorblox
Bull Case: $4B+ valuation; AI-native approach to email security; 15%+ of Fortune 500 as customers; behavioral AI detects zero-day attacks; $100M+ ARR
What Paul Graham Would Advise: Email security is boring but critical - that's your opportunity. Do things that don't scale: personally investigate attacks for your first customers. Remember, the best security products are invisible until they're needed. Be the hero that stops the breach.
Notable Investors: Greylock Partners, Menlo Ventures, Insight Partners, CrowdStrike, Slalom Build
Valuation: $5.0B | Funding: $1.0B | Location: Palo Alto, CA, USA
Team Size: ~600 employees | Category: AI Infrastructure
Product: AI hardware and software platform with custom chips (DataScale) for training and inference workloads
Key Competitors: NVIDIA, Groq, Cerebras, Graphcore, Intel Habana
Bull Case: $5B+ valuation; Stanford research background; custom AI chips; $100M+ revenue; enterprise focus; alternative to NVIDIA dependency
What Paul Graham Would Advise: Taking on NVIDIA is ambitious - that's good. Do things that don't scale: personally optimize workloads for your first customers. Remember, the best hardware companies win on software. Make your chips easy to use.
Notable Investors: Intel Capital, GV, Walden International, Atlantic Bridge, Redline Capital
Valuation: $5.0B | Funding: $0.8B | Location: Pittsburgh, PA, USA
Team Size: ~400 employees | Category: AI Healthcare
Product: AI-powered clinical documentation that automatically transcribes and summarizes medical conversations
Key Competitors: Nuance (Microsoft), Suki, Ambience Healthcare, DeepScribe, Nabla
Bull Case: $5.3B valuation; $300M Series E; 50+ health systems; reduces physician burnout; Carnegie Mellon roots; solving the $100B documentation problem
What Paul Graham Would Advise: Physician burnout is a real crisis - you're solving it. Do things that don't scale: personally shadow doctors to understand their pain. Remember, the best healthcare AI reduces administrative burden. Give doctors their time back.
Notable Investors: Redpoint Ventures, Lightspeed Venture Partners, Spark Capital, Bessemer Venture Partners
Valuation: $5.0B | Funding: $0.5B | Location: Singapore / San Francisco, CA, USA
Team Size: ~200 employees | Category: AI Backend
Product: Open-source Firebase alternative built on PostgreSQL with AI vector tools - default backend for AI apps
Key Competitors: Firebase, Hasura, PlanetScale, Neon, CockroachDB
Bull Case: $5B valuation; 1M+ developers; open-source approach; AI vector extensions; default backend for AI startups; capital efficient growth
What Paul Graham Would Advise: Open source is your moat - lean into it. Do things that don't scale: personally help developers build on Supabase. Remember, the best developer tools feel inevitable. Be the default choice for AI app backends.
Notable Investors: Coatue, Felicis Ventures, Lightspeed Venture Partners, Mozilla Ventures, Y Combinator
Valuation: $4.5B | Funding: $0.3B | Location: San Francisco, CA, USA
Team Size: ~100 employees | Category: AI Media
Product: Generative media platform for AI-powered image and video generation at scale
Key Competitors: Midjourney, Runway, Stability AI, DALL-E, Leonardo.ai
Bull Case: $4.5B valuation; $140M Series D; Sequoia backing; fastest generative media inference; powering AI apps that need media generation
What Paul Graham Would Advise: Speed is your moat in generative media. Do things that don't scale: personally optimize pipelines for your first customers. Remember, the best infrastructure is invisible. Make media generation so fast it feels instant.
Notable Investors: Sequoia Capital, Andreessen Horowitz, General Catalyst, Kindred Ventures
Valuation: $4.5B | Funding: $0.4B | Location: New York, NY, USA
Team Size: ~200 employees | Category: AI Platform
Product: ML model hub, Transformers library, and AI community platform - the GitHub of machine learning
Key Competitors: GitHub, Papers With Code, Kaggle, Replicate, Model Zoo
Bull Case: $4.5B valuation; 1M+ models hosted; 100K+ organizations; Transformers library is industry standard; community-driven network effects
What Paul Graham Would Advise: You've built the community that powers AI - that's incredibly valuable. Do things that don't scale: personally welcome new contributors. Remember, the best platforms serve their community first. Keep the open-source ethos alive.
Notable Investors: Salesforce Ventures, Google, Amazon, NVIDIA, Intel Capital, Lux Capital
Valuation: $4.0B | Funding: $2.0B | Location: Beijing, China
Team Size: ~300 employees | Category: Foundation Model
Product: Chinese LLM with ultra-long context windows (2M+ tokens) and strong reasoning capabilities
Key Competitors: OpenAI, Baidu (Ernie), Alibaba (Tongyi), 01.AI, Zhipu AI
Bull Case: China's leading AI lab; $3B+ funding; long context breakthroughs; Kimi chatbot popular in China; alternative to Western AI in Chinese market
What Paul Graham Would Advise: You're building for the largest market in the world. Do things that don't scale: personally onboard Chinese enterprises. Remember, the best products are culturally specific. Build AI that understands Chinese context.
Notable Investors: Alibaba, Sequoia Capital China, Hillhouse Capital, ZhenFund
Valuation: $4.0B | Funding: $2.0B | Location: Palo Alto, CA, USA
Team Size: ~62 employees | Category: Foundation Model
Product: Pi - Personal AI assistant focused on emotional intelligence and conversational warmth
Key Competitors: OpenAI (ChatGPT), Anthropic (Claude), Character.AI, Replika
Bull Case: $1.5B raised at $4B valuation; Mustafa Suleyman (DeepMind co-founder); emotional intelligence focus; acquired by Microsoft (March 2024)
What Paul Graham Would Advise: Emotional AI is underrated - lean into it. Do things that don't scale: personally chat with early users to understand what they need. Remember, the best AI feels like a friend. Make Pi someone people want to talk to.
Notable Investors: Microsoft, Greylock Partners, Dragoneer Investment Group
Valuation: $4.0B | Funding: $0.5B | Location: London, UK
Team Size: ~400 employees | Category: AI Video
Product: AI video generation platform that creates professional videos with AI avatars from text
Key Competitors: HeyGen, D-ID, Colossyan, Hour One, Elai
Bull Case: $4B valuation; 60K+ customers including 60% of Fortune 100; AI avatars in 130+ languages; $100M+ ARR; enterprise video at scale
What Paul Graham Would Advise: Video is the future of communication - you're making it accessible. Do things that don't scale: personally create videos for your first enterprise customers. Remember, the best products remove friction. Make video creation as easy as writing an email.
Notable Investors: Accel, Kleiner Perkins, GV, FirstMark Capital, MMC Ventures
Valuation: $4.0B | Funding: $1.0B | Location: Palo Alto, CA, USA
Team Size: ~150 employees | Category: AI Video/3D
Product: AI models for photo, video, and 3D generation - Dream Machine for video, Genie for 3D
Key Competitors: Runway, Pika Labs, Midjourney, Stability AI, Kaedim
Bull Case: $4B valuation; $900M Series C; fastest video generation; 3D capabilities; NeRF technology; enterprise and consumer applications
What Paul Graham Would Advise: 3D + video is the next frontier. Do things that don't scale: personally help creators make their first AI videos. Remember, the best creative tools inspire creativity. Make Luma feel like magic.
Notable Investors: Andreessen Horowitz, Amplify Partners, General Catalyst, Matrix Partners
Valuation: $4.0B | Funding: $0.5B | Location: San Francisco, CA, USA
Team Size: ~600 employees | Category: AI Security
Product: AI-powered security and compliance automation platform for SOC 2, ISO 27001, GDPR, HIPAA
Key Competitors: Drata, Secureframe, Tugboat Logic, Anecdotes
Bull Case: $4B valuation; 8,000+ customers; category creator for compliance automation; $100M+ ARR; essential for B2B SaaS startups
What Paul Graham Would Advise: Compliance is the schlep every startup avoids - that's your moat. Do things that don't scale: personally help your first customers get SOC 2. Remember, the best products turn pain into delight. Make compliance feel like a superpower.
Notable Investors: Sequoia Capital, Y Combinator, Craft Ventures, JPMorgan Growth Equity
Valuation: $3.7B | Funding: $1.0B | Location: Shanghai, China
Team Size: ~500 employees | Category: Foundation Model
Product: Chinese AI platform with text, voice, and video generation capabilities - Hailuo AI video generator
Key Competitors: Moonshot AI, 01.AI, Zhipu AI, Baidu Ernie, Alibaba Tongyi
Bull Case: $1B+ funding; leading Chinese AI startup; multi-modal capabilities; Hailuo AI popular for video generation; strong in Chinese market
What Paul Graham Would Advise: Multi-modal AI is the future - you're ahead of the curve. Do things that don't scale: personally help Chinese creators use your tools. Remember, the best products work seamlessly across modalities. Make AI that understands all forms of content.
Notable Investors: Tencent, Alibaba, Hillhouse Capital, Qiming Venture Partners
Valuation: $3.3B | Funding: $0.5B | Location: San Francisco, CA, USA
Team Size: ~287 employees | Category: AI Infrastructure
Product: Decentralized cloud platform for training and deploying open-source AI models with competitive pricing
Key Competitors: CoreWeave, Lambda Labs, Fireworks AI, Replicate, Baseten
Bull Case: $3.3B valuation; $305M Series B; open-source model focus; $44M revenue; 50% cheaper than alternatives; Sequoia and NVIDIA backing
What Paul Graham Would Advise: Open source is your moat - lean into it. Do things that don't scale: personally optimize workloads for your first customers. Remember, the best infrastructure companies win on price and performance. Be the AWS for open-source AI.
Notable Investors: Sequoia Capital, General Catalyst, Prosperity7, NVIDIA, Kleiner Perkins
Valuation: $3.3B | Funding: $0.3B | Location: New York, NY, USA
Team Size: ~400 employees | Category: AI Audio
Product: AI voice technology - text-to-speech, voice cloning, dubbing in 32 languages with lifelike quality
Key Competitors: Descript, WellSaid Labs, Murf.ai, Play.ht, Resemble AI
Bull Case: $6.6B valuation; $330M ARR; 41% of Fortune 500 as customers; most realistic AI voices; powering voice agents and content at scale
What Paul Graham Would Advise: Voice is the most natural interface - you're making it accessible. Do things that don't scale: personally clone voices for your first customers. Remember, the best AI voice feels human. Make people forget they're talking to AI.
Notable Investors: Andreessen Horowitz, ICONIQ Growth, Sequoia Capital, Salesforce Ventures, SV Angel
Valuation: $3.2B | Funding: $0.5B | Location: San Francisco, CA, USA
Team Size: ~80 employees | Category: Foundation Model
Product: Generative AI for images and video - Flux.1 model powers xAI Grok's image generation
Key Competitors: Midjourney, Stability AI, OpenAI (DALL-E), Adobe Firefly
Bull Case: $3.3B valuation; $300M Series B; former Stability AI team; Flux.1 open-source model; powers xAI Grok; rapid commercial traction
What Paul Graham Would Advise: You've built one of the best image models in months - that's execution. Do things that don't scale: personally help creators use Flux. Remember, the best generative AI inspires creativity. Make Black Forest the default for image generation.
Notable Investors: Andreessen Horowitz, Lightspeed Venture Partners, General Catalyst
Valuation: $3.2B | Funding: $0.4B | Location: San Francisco, CA, USA
Team Size: ~823 employees | Category: AI DevTools
Product: Frontend cloud platform; creator of Next.js; AI SDK for building AI apps; v0 AI UI generator
Key Competitors: Netlify, AWS Amplify, Cloudflare Pages, Gatsby, Deno Deploy
Bull Case: $9.3B valuation; powers 4M+ websites; $200M ARR; Next.js is dominant React framework; AI SDK gets 3M weekly downloads; default for AI web apps
What Paul Graham Would Advise: You've built the default frontend platform - now own AI apps. Do things that don't scale: personally help developers build their first AI app. Remember, the best platforms enable others to build. Be the infrastructure for the AI web.
Notable Investors: Accel, GIC, Bedrock Capital, GV, 8VC
Valuation: $3.2B | Funding: $0.3B | Location: San Francisco, CA, USA
Team Size: ~500 employees | Category: AI DevTools
Product: Low-code platform for building internal tools with AI-powered features for rapid development
Key Competitors: OutSystems, Mendix, Bubble, Appsmith, ToolJet
Bull Case: $3.2B valuation; 10,000+ customers including Amazon, Stripe, Volvo; $100M+ ARR; internal tools are the schlep everyone avoids
What Paul Graham Would Advise: Internal tools are the ultimate schlep - that's your opportunity. Do things that don't scale: personally build tools for your first customers. Remember, the best products solve real workflow pain. Make internal tool building feel like magic.
Notable Investors: Sequoia Capital, Stripe, John Collison, Nat Friedman, Elad Gil
Valuation: $3.1B | Funding: $0.1B | Location: San Francisco, CA, USA
Team Size: ~50 employees | Category: AI Research
Product: Real-time generative AI platform for instant AI-generated content and experiences
Key Competitors: OpenAI, Midjourney, Runway, Fal, Stability AI
Bull Case: $3.1B valuation; $100M+ raised; Benchmark and Sequoia backing; real-time generation; sub-second AI content creation
What Paul Graham Would Advise: Speed is everything in generative AI. Do things that don't scale: personally optimize latency for your first customers. Remember, the best AI feels instant. Make generation so fast it feels like thought.
Notable Investors: Benchmark, Sequoia Capital, Thrive Capital, General Catalyst
Valuation: $3.0B | Funding: $0.5B | Location: Beijing, China
Team Size: ~400 employees | Category: Foundation Model
Product: Chinese LLM platform with ChatGLM models - academic-focused with strong research capabilities
Key Competitors: Moonshot AI, MiniMax, 01.AI, Baidu Ernie, Alibaba Tongyi
Bull Case: $3B valuation; Tsinghua University roots; academic credibility; ChatGLM open-source models; strong technical team
What Paul Graham Would Advise: Academic credibility is your moat in China. Do things that don't scale: personally work with researchers using your models. Remember, the best AI comes from the best research. Bridge academia and industry.
Notable Investors: Alibaba, Hillhouse Capital, Shunwei Capital, Qiming Venture Partners
Valuation: $3.0B | Funding: $2.0B | Location: San Francisco, CA, USA
Team Size: ~400 employees | Category: AI Infrastructure
Product: Deep learning infrastructure and GPU cloud for AI training and inference workloads
Key Competitors: CoreWeave, Together AI, AWS, Google Cloud, Paperspace
Bull Case: $2.5B valuation; $900M+ raised; $480M Series D; Microsoft partnership; GPU cloud for AI; cheaper than hyperscalers
What Paul Graham Would Advise: GPU infrastructure is the picks and shovels of AI. Do things that don't scale: personally optimize clusters for your first customers. Remember, the best infrastructure companies win on reliability. Be the cloud AI engineers trust.
Notable Investors: Andreessen Horowitz, Coatue, Tiger Global, 1517 Fund, Bloomberg Beta
Valuation: $3.0B | Funding: $0.4B | Location: Tokyo, Japan
Team Size: ~50 employees | Category: AI Infrastructure
Product: AI model development using nature-inspired approaches - evolutionary algorithms for model creation
Key Competitors: OpenAI, Anthropic, Cohere, Mistral AI, Stability AI
Bull Case: $3B valuation; Japan's AI champion; former Google Brain researchers; unique approach to model development; government backing
What Paul Graham Would Advise: Nature-inspired AI is unique - lean into it. Do things that don't scale: personally evolve models for your first use cases. Remember, the best research comes from first principles. Question how everyone else builds AI.
Notable Investors: NTT Group, KDDI, Sony, NVIDIA, Lux Capital
Valuation: $3.0B | Funding: $0.5B | Location: New York, NY, USA
Team Size: ~382 employees | Category: AI Video
Product: AI-powered creative tools for video generation, editing, and visual effects - Gen-4 video model
Key Competitors: Midjourney, Pika Labs, Luma AI, Stability AI, Kling AI
Bull Case: $3B valuation; $536M raised; Hollywood partnerships; $300M revenue target; leading AI video generation; creative professional focus
What Paul Graham Would Advise: Video is the future of creativity - you're making it accessible. Do things that don't scale: personally help filmmakers use Runway. Remember, the best creative tools inspire. Make Runway feel like a superpower for creators.
Notable Investors: General Atlantic, Felicis Ventures, Amplify Partners, Lux Capital, Madrona Venture Group
Valuation: $3.0B | Funding: $0.5B | Location: Paris, France
Team Size: ~100 employees | Category: AI Coding
Product: AI coding platform with proprietary models designed specifically for code generation
Key Competitors: GitHub Copilot, Cursor, Cognition AI, Magic, Codeium
Bull Case: $3B valuation; $626M raised; $500M Series B; Bain Capital led; purpose-built for code; European AI champion
What Paul Graham Would Advise: Purpose-built models for code is smart - lean into it. Do things that don't scale: personally pair program with early users. Remember, the best coding AI understands intent, not just syntax. Make Poolside feel like a brilliant pair programmer.
Notable Investors: Bain Capital, eBay, NVIDIA, DST Global, Redpoint Ventures
Valuation: $2.9B | Funding: $0.2B | Location: Mountain View, CA, USA
Team Size: ~150 employees | Category: AI Coding
Product: AI-powered coding platform with Windsurf IDE - free alternative to GitHub Copilot
Key Competitors: GitHub Copilot, Cursor, Amazon CodeWhisperer, Tabnine
Bull Case: $2.9B valuation; $150M raised; free tier drives adoption; Windsurf IDE acquisition; 1M+ developers; acquired by OpenAI for $3B
What Paul Graham Would Advise: Free is a powerful acquisition strategy. Do things that don't scale: personally help developers set up Codeium. Remember, the best developer tools spread virally. Make Codeium so good developers can't help but recommend it.
Notable Investors: Kleiner Perkins, General Catalyst, Greenoaks, Elad Gil, Nat Friedman
Valuation: $2.7B | Funding: $0.7B | Location: Toronto, Canada / Santa Clara, CA, USA
Team Size: ~600 employees | Category: AI Infrastructure
Product: AI hardware company building processors and systems for AI training and inference
Key Competitors: NVIDIA, AMD, Intel, Cerebras, SambaNova, Groq
Bull Case: $2.7B valuation; $693M raised; Jim Keller (chip legend) leading; open-source software stack; alternative to NVIDIA monopoly
What Paul Graham Would Advise: Taking on NVIDIA requires world-class talent - you have it. Do things that don't scale: personally work with early adopters on chip design. Remember, the best hardware companies win on software. Make Tenstorrent easy to program.
Notable Investors: Fidelity, Hyundai, Samsung, Eclipse Ventures, Real Ventures
Valuation: $2.5B | Funding: $0.5B | Location: Santa Clara, CA, USA
Team Size: ~200 employees | Category: AI Infrastructure
Product: Optical interconnect technology for AI data centers - Photonic Fabric for faster AI training
Key Competitors: NVIDIA, AMD, Intel, Ayar Labs, Lightmatter
Bull Case: $2.5B valuation; $250M+ raised; optical interconnect breakthrough; solves data movement bottleneck; strategic for AI infrastructure
What Paul Graham Would Advise: Optical interconnect is the future of AI infrastructure. Do things that don't scale: personally design systems for your first customers. Remember, the best infrastructure innovations are invisible. Make data movement so fast it disappears.
Notable Investors: Tiger Global, Koch Disruptive Technologies, Temasek, Samsung, Porsche
Valuation: $2.5B | Funding: $0.3B | Location: Cambridge, MA, USA
Team Size: ~100 employees | Category: AI Audio
Product: AI music generation platform that creates songs from text prompts in any style
Key Competitors: Udio, AIVA, Boomy, Soundraw, Loudly
Bull Case: $2.5B valuation; $250M Series C; 100M users; 7M tracks created daily; Menlo Ventures backing; democratizing music creation
What Paul Graham Would Advise: Music is universal - you're democratizing it. Do things that don't scale: personally help creators make their first AI song. Remember, the best creative AI inspires. Make Suno feel like having a band in your pocket.
Notable Investors: Menlo Ventures, Lightspeed Venture Partners, Matrix Partners, Founder Collective
Valuation: $2.4B | Funding: $0.6B | Location: South San Francisco, CA, USA
Team Size: ~300 employees | Category: AI Drug Discovery
Product: AI-powered drug discovery platform using machine learning to predict molecular behavior
Key Competitors: Recursion, Exscientia, Isomorphic Labs, Atomwise, BenevolentAI
Bull Case: $2.4B valuation; Daphne Koller (Coursera co-founder) leading; $600M+ raised; partnerships with pharma giants; AI + biology expertise
What Paul Graham Would Advise: Drug discovery is the ultimate schlep - that's why AI is needed. Do things that don't scale: personally work with biologists on your first targets. Remember, the best AI drug companies combine ML and biology expertise. Hire both.
Notable Investors: Andreessen Horowitz, GV, Foresite Capital, Casdin Capital, Third Rock Ventures
Valuation: $2.2B | Funding: $0.2B | Location: Palo Alto, CA, USA
Team Size: ~1,000 employees | Category: AI Talent
Product: AI-powered talent platform connecting companies with vetted remote developers worldwide
Key Competitors: Toptal, Andela, Gun.io, Arc.dev, Lemon.io
Bull Case: $2.2B valuation; $200M raised; 3M+ developers in network; AI matching; used by Google, Meta, Netflix; remote work tailwinds
What Paul Graham Would Advise: Talent is the bottleneck for every company - you're solving it. Do things that don't scale: personally interview your first 100 developers. Remember, the best marketplaces have high-quality supply. Vet every developer like your reputation depends on it.
Notable Investors: Westbridge Capital, Foundation Capital, Altos Ventures, Gaingels
Valuation: $2.2B | Funding: $0.4B | Location: New York, NY, USA
Team Size: ~300 employees | Category: AI Automation
Product: AI-powered automation platform for property management - leasing, maintenance, resident communication
Key Competitors: AppFolio, Buildium, Yardi, Entrata, RealPage
Bull Case: $2.2B valuation; $400M raised; 70%+ of top property managers as customers; $100M+ ARR; AI-native approach to real estate
What Paul Graham Would Advise: Property management is the ultimate schlep - perfect for AI. Do things that don't scale: personally handle leasing for your first properties. Remember, the best vertical AI solves real workflow pain. Make property managers' lives easier.
Notable Investors: GV, Navitas Capital, AvalonBay Communities, Cushman & Wakefield
Valuation: $2.1B | Funding: $0.3B | Location: Mountain View, CA, USA
Team Size: ~733 employees | Category: AI Support
Product: AI-powered employee support platform for IT, HR, and workplace operations automation
Key Competitors: Espressive, BMC Software, ServiceNow, Ada, Forethought
Bull Case: $2.1B valuation; $308M raised; $100M revenue; acquired for $2.85B (March 2025); enterprise automation at scale; IT chatbot leader
What Paul Graham Would Advise: IT support is the schlep everyone hates - that's your opportunity. Do things that don't scale: personally resolve tickets for your first customers. Remember, the best AI support feels like magic. Make Moveworks the IT person everyone wishes they had.
Notable Investors: Tiger Global, Bain Capital Ventures, Lightspeed Venture Partners, Sapphire Ventures
Valuation: $2.0B | Funding: $0.1B | Location: Cologne, Germany
Team Size: ~800 employees | Category: AI Translation
Product: AI-powered translation platform with superior quality to Google Translate for professional use
Key Competitors: Google Translate, Microsoft Translator, Amazon Translate, Smartcat, Phrase
Bull Case: $2B valuation; $100M raised; bootstrapped to $100M+ ARR; best-in-class translation quality; 100K+ business customers; European AI success
What Paul Graham Would Advise: Translation quality is your moat - protect it. Do things that don't scale: personally translate for your first enterprise customers. Remember, the best products win on quality. Be so good that professionals can't ignore you.
Notable Investors: IVP, Atomico, Bessemer Venture Partners, Benchmark
Valuation: $1.0B | Funding: $0.2B | Location: Menlo Park, CA, USA
Team Size: ~225 employees | Category: Foundation Model
Product: Conversational AI platform for creating and interacting with AI characters and personalities
Key Competitors: Replika, Hyro AI, Inworld AI, Soul Machines
Bull Case: $1B valuation; $193M raised; 20M+ monthly users; strong engagement; entertainment-focused AI; Google acquisition discussions
What Paul Graham Would Advise: Character AI is the future of entertainment - lean into it. Do things that don't scale: personally create characters with your first users. Remember, the best AI feels like a friend. Make Character.ai where people go when they want to chat.
Notable Investors: Andreessen Horowitz, Nat Friedman, Elad Gil, SV Angel
Valuation: $2.0B | Funding: $0.3B | Location: Boston, MA, USA
Team Size: ~100 employees | Category: Foundation Model
Product: Foundation models using liquid neural networks - more efficient and interpretable than transformers
Key Competitors: OpenAI, Anthropic, Mistral AI, Cohere, Reka AI
Bull Case: $2B valuation; $250M+ raised; MIT research background; liquid neural networks; more efficient than transformers; unique technical approach
What Paul Graham Would Advise: New architectures are risky but potentially transformative. Do things that don't scale: personally benchmark your models against transformers. Remember, the best research challenges assumptions. Prove liquid networks are the future.
Notable Investors: AMD, DJI, WordPress.com, Samsung Next, BMW i Ventures
Valuation: $1.0B | Funding: $0.1B | Location: San Francisco, CA, USA
Team Size: ~50 employees | Category: Foundation Model
Product: Enterprise-focused multimodal AI models for text, image, and video understanding
Key Competitors: OpenAI, Anthropic, Cohere, AI21 Labs
Bull Case: $1B valuation; $100M+ raised; DeepMind alumni; multimodal focus; enterprise applications; efficient model training
What Paul Graham Would Advise: Multimodal is the future - you're positioned well. Do things that don't scale: personally deploy models for your first enterprise customers. Remember, the best enterprise AI solves real business problems. Make Reka the obvious choice for multimodal.
Notable Investors: DST Global, Snowflake Ventures, Radical Ventures, Lux Capital
Valuation: $1.0B | Funding: $0.0B | Location: Beijing, China
Team Size: ~200 employees | Category: Foundation Model
Product: Yi series of open-source LLMs - China's answer to Llama with strong performance
Key Competitors: Moonshot AI, MiniMax, Zhipu AI, Baidu Ernie, Alibaba Tongyi
Bull Case: $1B valuation; Kai-Fu Lee (AI legend) leading; open-source approach; strong Chinese market position; zero external funding shows capital efficiency
What Paul Graham Would Advise: Open source in China is bold - lean into it. Do things that don't scale: personally help developers use Yi models. Remember, the best open-source projects have passionate communities. Build the Chinese AI community around 01.AI.
Notable Investors: Self-funded / Kai-Fu Lee
Valuation: $2.0B | Funding: $0.2B | Location: Santa Clara, CA, USA
Team Size: ~150 employees | Category: AI Infrastructure
Product: AI inference chips using in-memory computing for efficient transformer processing
Key Competitors: NVIDIA, Groq, Cerebras, SambaNova, Tenstorrent
Bull Case: $2B valuation; $200M+ raised; in-memory computing breakthrough; inference-focused; Microsoft partnership; efficient transformer processing
What Paul Graham Would Advise: Inference is the next battleground - you're positioned well. Do things that don't scale: personally optimize inference for your first customers. Remember, the best chip companies win on efficiency. Make d-Matrix the obvious choice for inference.
Notable Investors: Microsoft, Marvell, Playground Global, M12, Nautilus Venture Partners
Valuation: $2.0B | Funding: $0.1B | Location: San Francisco, CA, USA
Team Size: ~100 employees | Category: AI Productivity
Product: AI-powered presentation and document creation platform - alternative to PowerPoint and Google Slides
Key Competitors: Beautiful.ai, Tome, Pitch, Canva, Google Slides
Bull Case: $2B valuation; $100M+ raised; 20M+ users; AI-first approach to presentations; viral growth; modern alternative to PowerPoint
What Paul Graham Would Advise: Presentations are the schlep everyone hates - that's your opportunity. Do things that don't scale: personally design decks for your first customers. Remember, the best productivity tools save time. Make Gamma feel like having a designer on demand.
Notable Investors: Accel, General Catalyst, GV, Liquid 2 Ventures
Valuation: $2.0B | Funding: $0.4B | Location: Mountain View, CA, USA
Team Size: ~600 employees | Category: AI HR
Product: AI-powered talent intelligence platform for recruiting, retention, and workforce planning
Key Competitors: LinkedIn Talent Solutions, Workday, Greenhouse, Lever, Phenom
Bull Case: $2B+ valuation; $400M+ raised; 100+ enterprise customers; AI matching for talent; former Google Search team; $100M+ ARR
What Paul Graham Would Advise: Talent is the most important thing in every company - you're optimizing it. Do things that don't scale: personally match candidates for your first customers. Remember, the best HR AI makes great hires. Be the reason companies find their best people.
Notable Investors: General Catalyst, IVP, Foundation Capital, Lightspeed Venture Partners
Valuation: $2.0B | Funding: $0.3B | Location: London, UK
Team Size: ~200 employees | Category: AI Drug Discovery
Product: AI-powered drug discovery using AlphaFold technology to predict protein structures
Key Competitors: Insitro, Recursion, Exscientia, Atomwise, BenevolentAI
Bull Case: $2B+ valuation; DeepMind spinout; AlphaFold technology; Demis Hassabis leading; potential to revolutionize drug discovery
What Paul Graham Would Advise: You're applying the most important AI breakthrough to drug discovery. Do things that don't scale: personally work with pharma partners on first targets. Remember, the best AI drug companies have deep scientific expertise. Bridge AI and biology.
Notable Investors: Alphabet, Thrive Capital, Andreessen Horowitz, GV
Valuation: $2.0B | Funding: $0.5B | Location: Salt Lake City, UT, USA
Team Size: ~500 employees | Category: AI Drug Discovery
Product: AI-powered drug discovery platform using automated experiments and machine learning
Key Competitors: Insitro, Exscientia, Isomorphic Labs, Atomwise, BenevolentAI
Bull Case: $2B+ valuation; $500M+ raised; public company (RXRX); massive experimental dataset; Nvidia partnership; AI + wet lab integration
What Paul Graham Would Advise: Data is your moat in AI drug discovery - keep generating it. Do things that don't scale: personally analyze experiments with your team. Remember, the best AI drug companies learn from every molecule. Make each failure teach you something.
Notable Investors: Baillie Gifford, Nvidia, Obvious Ventures, Lux Capital, Mubadala
Valuation: $1.2B | Funding: $0.3B | Location: Tel Aviv, Israel
Team Size: ~200 employees | Category: AI Infrastructure
Product: Edge AI processors for running AI on devices - Hailo-8 chip for low-power inference
Key Competitors: NVIDIA Jetson, Intel Movidius, Qualcomm, Ambarella, EdgeQ
Bull Case: $1.2B valuation; $300M+ raised; edge AI focus; 26 TOPS at low power; automotive and industrial applications; Israeli tech excellence
What Paul Graham Would Advise: Edge AI is the future - you're positioned well. Do things that don't scale: personally optimize models for your first edge deployments. Remember, the best edge AI is invisible. Make Hailo the obvious choice for on-device AI.
Notable Investors: Poalim Equity, OurCrowd, Millennium Technology Value Partners, NEC Corporation
Valuation: $1.4B | Funding: $0.3B | Location: Seoul, South Korea
Team Size: ~150 employees | Category: AI Infrastructure
Product: AI chips for data centers and edge devices - Atom and Ion processors for inference
Key Competitors: NVIDIA, Samsung, SK Hynix, Tenstorrent, SambaNova
Bull Case: $1.4B valuation; $300M+ raised; South Korea's AI chip champion; government backing; strategic for Korean tech independence
What Paul Graham Would Advise: National champions have unique advantages - leverage yours. Do things that don't scale: personally work with Korean tech giants. Remember, the best chip companies have deep partnerships. Make Rebellions the pride of Korean AI.
Notable Investors: Korea Development Bank, SK Telecom, KB Investment, Korea Investment Partners
Valuation: $1.5B | Funding: $0.3B | Location: Pittsburgh, PA, USA
Team Size: ~100 employees | Category: AI Robotics
Product: General-purpose AI models for robotics - brain that can power any robot
Key Competitors: Physical Intelligence, Covariant, Sanctuary AI, Figure, Tesla Optimus
Bull Case: $1.5B valuation; $300M Series A; Carnegie Mellon roots; general-purpose robot brain; massive market potential
What Paul Graham Would Advise: General-purpose robotics is the holy grail - go for it. Do things that don't scale: personally train robots for your first customers. Remember, the best robotics companies collect the best data. Every robot is a data collection opportunity.
Notable Investors: Lightspeed Venture Partners, Coatue, Jeff Bezos, SoftBank, Felicis Ventures
Valuation: $1.5B | Funding: $0.1B | Location: Austin, TX, USA
Team Size: ~300 employees | Category: AI Writing
Product: AI content writing platform for marketing copy, blog posts, and social media
Key Competitors: Copy.ai, Writer, Grammarly, Anyword, Hypotenuse AI
Bull Case: $1.5B valuation; early mover in AI copywriting; 100K+ customers; brand recognition; marketing focus
What Paul Graham Would Advise: You were early - now stay relevant. Do things that don't scale: personally write copy for your first enterprise customers. Remember, the best AI writing understands brand voice. Make Jasper feel like an extension of marketing teams.
Notable Investors: Insight Partners, Coatue, Bessemer Venture Partners, IVP, HubSpot Ventures
Valuation: $1.5B | Funding: $0.2B | Location: Palo Alto, CA, USA
Team Size: ~100 employees | Category: AI Search
Product: AI-powered search engine with privacy focus and customizable results
Key Competitors: Google, Perplexity, Bing, DuckDuckGo, Brave Search
Bull Case: $1.5B valuation; $200M+ raised; privacy-focused; AI summaries; customizable search; ex-Salesforce CEO leading
What Paul Graham Would Advise: Privacy + AI is a powerful combination. Do things that don't scale: personally help users customize their search. Remember, the best search feels like it knows you. Make You.com the search engine people trust.
Notable Investors: Salesforce Ventures, Norwest Venture Partners, IVP, Radical Ventures
Valuation: $1.0B | Funding: $0.2B | Location: San Francisco, CA, USA
Team Size: ~200 employees | Category: AI Healthcare
Product: AI-powered clinical documentation and workflow automation for healthcare providers
Key Competitors: Abridge, Nuance, Suki, DeepScribe, Nabla
Bull Case: $1B+ valuation; $243M raised; healthcare OS approach; multiple specialties; reducing physician burnout
What Paul Graham Would Advise: Healthcare workflows are broken - you're fixing them. Do things that don't scale: personally shadow physicians in different specialties. Remember, the best healthcare AI reduces administrative burden. Give doctors their time back.
Notable Investors: Kleiner Perkins, OpenAI Startup Fund, Andreessen Horowitz, Optum Ventures
Valuation: $1.0B | Funding: $1.0B | Location: San Francisco, CA, USA
Team Size: ~150 employees | Category: AI Drug Discovery
Product: AI-powered drug discovery platform using generative AI to design new therapeutics
Key Competitors: Insitro, Recursion, Isomorphic Labs, Exscientia, BenevolentAI
Bull Case: $1B+ valuation; $1B raised; Flagship Pioneering backing; generative AI for drug design; top scientific team
What Paul Graham Would Advise: Generative AI for drugs is the ultimate application. Do things that don't scale: personally work with scientists on first molecules. Remember, the best AI drug companies combine AI expertise with scientific rigor. Bridge both worlds.
Notable Investors: Flagship Pioneering, F-Prime Capital, ARCH Venture Partners, GV
Valuation: $1.8B | Funding: $0.2B | Location: Stockholm, Sweden
Team Size: ~50 employees | Category: AI Development
Product: AI app builder that creates full applications from text prompts - 'vibe coding' platform
Key Competitors: Replit, Vercel v0, Bolt, Softr, Bubble
Bull Case: $6.6B valuation; $330M raised; $200M ARR; Europe's leading 'vibe coding' platform; 100K projects daily; fastest growth in Europe
What Paul Graham Would Advise: Full-stack AI generation is the future - you're leading it. Do things that don't scale: personally help users build their first apps. Remember, the best no-code tools empower creators. Make Lovable feel like having a dev team on demand.
Notable Investors: CapitalG, Menlo Ventures, EQT Ventures, Creandum
Valuation: $1.0B | Funding: $0.1B | Location: San Diego, CA, USA
Team Size: ~100 employees | Category: AI Drug Discovery
Product: AI-powered drug discovery platform for small molecule therapeutics
Key Competitors: Insitro, Recursion, Exscientia, Isomorphic Labs, Atomwise
Bull Case: $1B+ valuation; $100M+ raised; AI-driven chemistry; multiple programs in development; top scientific team
What Paul Graham Would Advise: Small molecules are the perfect AI drug target. Do things that don't scale: personally work with chemists on first compounds. Remember, the best AI drug companies move fast from prediction to validation. Speed is your advantage.
Notable Investors: Nexus Venture Partners, Coatue, Foresite Capital, Casdin Capital
Valuation: $1.0B | Funding: $0.2B | Location: Burlingame, CA, USA
Team Size: ~80 employees | Category: AI Drug Discovery
Product: AI-powered drug discovery using generative models for molecular design
Key Competitors: Insitro, Recursion, Xaira, Iambic, Exscientia
Bull Case: $1B+ valuation; $200M+ raised; Stanford roots; generative molecular design; partnerships with pharma
What Paul Graham Would Advise: Generative molecular design is cutting-edge - stay ahead. Do things that don't scale: personally design molecules with your team. Remember, the best AI drug companies validate quickly. Move from computer to lab fast.
Notable Investors: Andreessen Horowitz, Fidelity, Radical Ventures, T. Rowe Price
Valuation: $1.5B | Funding: $0.3B | Location: Oxford, UK
Team Size: ~400 employees | Category: AI Drug Discovery
Product: AI-powered drug discovery platform with end-to-end capabilities from design to clinic
Key Competitors: Insitro, Recursion, Isomorphic Labs, BenevolentAI, Atomwise
Bull Case: $1.5B valuation; public company (EXAI); first AI-designed drug in clinical trials; end-to-end platform; partnerships with pharma
What Paul Graham Would Advise: End-to-end drug discovery is ambitious - that's your moat. Do things that don't scale: personally work on every program. Remember, the best AI drug companies learn from every molecule. Make each failure teach you something.
Notable Investors: SoftBank Vision Fund, Oxford Science Enterprises, GT Healthcare Capital
Valuation: $1.0B | Funding: $0.0B | Location: Memphis, TN, USA
Team Size: ~100 employees | Category: AI Writing
Product: AI-powered copywriting and go-to-market platform for marketing teams
Key Competitors: Jasper, Writer, Anyword, Hypotenuse AI, Copysmith
Bull Case: $1B+ valuation; bootstrapped to $50M+ ARR; product-led growth; 5K+ customers; marketing workflow focus
What Paul Graham Would Advise: Bootstrapped to unicorn status - that's incredible. Do things that don't scale: personally write copy for your first customers. Remember, the best products spread through word of mouth. Make Copy.ai so good marketers can't stop talking about it.
Notable Investors: Self-funded / Bootstrapped
Valuation: $1.0B | Funding: $0.1B | Location: San Francisco, CA, USA
Team Size: ~150 employees | Category: AI Education
Product: AI-powered language learning app with personalized tutoring and conversation practice
Key Competitors: Duolingo, Babbel, Busuu, Pimsleur, Rosetta Stone
Bull Case: $1B+ valuation; $100M+ raised; AI tutoring; OpenAI partnership; 10M+ users; personalized learning at scale
What Paul Graham Would Advise: Language learning is perfect for AI - personalized, conversational, scalable. Do things that don't scale: personally tutor your first users. Remember, the best education AI makes learning feel effortless. Make Speak feel like having a personal tutor.
Notable Investors: OpenAI Startup Fund, Lachy Groom, Founders Fund, Emergence Capital
Valuation: $1.0B | Funding: $0.1B | Location: Washington, DC, USA
Team Size: ~50 employees | Category: AI Data
Product: Open-source platform for extracting and processing unstructured data for LLM applications
Key Competitors: LlamaIndex, LangChain, AWS Textract, Google Document AI
Bull Case: $1B+ valuation; $100M+ raised; open-source adoption; essential for RAG; data preprocessing for LLMs
What Paul Graham Would Advise: Data preprocessing is the schlep every AI app needs - that's your moat. Do things that don't scale: personally extract data for your first customers. Remember, the best infrastructure is invisible. Make Unstructured the default for data prep.
Notable Investors: Madrona Venture Group, Bain Capital Ventures, M12, Mango Capital
Valuation: $1.0B | Funding: $0.2B | Location: San Francisco, CA, USA
Team Size: ~302 employees | Category: AI MLOps
Product: MLOps platform for experiment tracking, model management, and collaboration
Key Competitors: MLflow, Neptune, Comet, ClearML, TensorBoard
Bull Case: $1B+ valuation; $250M raised; industry standard for ML tracking; acquired by CoreWeave for $1.7B; 500K+ users
What Paul Graham Would Advise: You've built the standard tool for ML engineers - that's powerful. Do things that don't scale: personally help teams set up experiment tracking. Remember, the best developer tools become indispensable. Make W&B the default for ML teams.
Notable Investors: Felicis Ventures, Insight Partners, Bond, Coatue, Lightspeed Venture Partners
Valuation: $1.25B | Funding: $0.26B | Location: San Francisco, CA, USA
Team Size: ~163 employees | Category: AI Developer Tools
Product: Framework for building LLM applications with chains, agents, and RAG - LangSmith for observability
Key Competitors: LlamaIndex, FlowiseAI, Langdock, Vercel AI SDK
Bull Case: $1.25B valuation; $260M raised; most popular LLM framework; 118K GitHub stars; LangGraph for orchestration; essential for AI apps
What Paul Graham Would Advise: You've built the default framework for LLM apps - that's incredible. Do things that don't scale: personally help developers build their first chains. Remember, the best frameworks enable others to build. Be the foundation for the AI app ecosystem.
Notable Investors: Benchmark, Sequoia Capital, IVP, Gradient Ventures, Felicis Ventures
Valuation: $7.25B | Funding: $0.77B | Location: Palo Alto, CA, USA
Team Size: ~1,000 employees | Category: AI Search
Product: AI-powered enterprise search and knowledge management platform - search across all company apps
Key Competitors: Coveo, Elastic, Algolia, Microsoft Copilot, Guru
Bull Case: $7.25B valuation; $765M raised; 100+ enterprise app integrations; $100M+ ARR; workplace knowledge discovery; ex-Google Search team
What Paul Graham Would Advise: Enterprise search is the schlep everyone needs - that's your opportunity. Do things that don't scale: personally index data for your first customers. Remember, the best enterprise AI makes knowledge accessible. Make Glean the company brain.
Notable Investors: Sequoia Capital, Kleiner Perkins, Lightspeed Venture Partners, IVP, General Catalyst
Valuation: $10.5B | Funding: $0.15B | Location: San Francisco, CA, USA
Team Size: ~131 employees | Category: AI Media
Product: AI image generation platform via Discord - highest quality artistic image generation
Key Competitors: DALL-E, Stable Diffusion, Adobe Firefly, Leonardo.ai, Ideogram
Bull Case: $10.5B valuation; $500M revenue; $0 marketing spend; 16.4M+ users; bootstrapped; highest quality AI images; lean team model
What Paul Graham Would Advise: Bootstrapped to $500M revenue - you're the dream. Do things that don't scale: personally engage with your Discord community. Remember, the best products spread through love, not ads. Keep Midjourney weird and wonderful.
Notable Investors: Lightspeed Venture Partners (first external round)
Valuation: $0.75B | Funding: $0.14B | Location: San Francisco, CA, USA
Team Size: ~133 employees | Category: AI Infrastructure
Product: Vector database platform for similarity search and RAG applications - managed vector search
Key Competitors: Weaviate, Chroma, Qdrant, Zilliz, LanceDB
Bull Case: $750M valuation; $138M raised; created vector DB category; essential for RAG; 10K+ customers; a16z backing
What Paul Graham Would Advise: You created the vector database category - now own it. Do things that don't scale: personally help customers implement RAG. Remember, the best infrastructure companies educate their market. Make Pinecone synonymous with vector search.
Notable Investors: Andreessen Horowitz, Menlo Ventures, Wing Venture Capital, Tiger Global
Valuation: $0.5B | Funding: $0.07B | Location: Amsterdam, Netherlands
Team Size: ~104 employees | Category: AI Infrastructure
Product: Open-source AI-native vector database with hybrid search capabilities
Key Competitors: Pinecone, Chroma, Qdrant, Zilliz, Vespa
Bull Case: $500M+ valuation; $67M raised; open-source approach; European HQ; hybrid search; 50K+ developers; Index Ventures backing
What Paul Graham Would Advise: Open source + European is a unique position - lean into it. Do things that don't scale: personally help developers set up Weaviate. Remember, the best open-source projects have passionate communities. Build the European AI infrastructure community.
Notable Investors: Index Ventures, Battery Ventures, Newion, Zetta Venture Partners
Valuation: $6.9B | Funding: $0.75B | Location: Mountain View, CA, USA
Team Size: ~400 employees | Category: AI Infrastructure
Product: LPU (Language Processing Unit) chips for ultra-fast AI inference - 750 tokens/sec
Key Competitors: NVIDIA, SambaNova, Cerebras, d-Matrix, Tenstorrent
Bull Case: $6.9B valuation; $750M raised; 750 tokens/sec inference; Jonathan Ross (Google TPU inventor); acquired by NVIDIA for $20B; fastest inference
What Paul Graham Would Advise: Speed is everything in inference - you're the fastest. Do things that don't scale: personally optimize workloads for your first customers. Remember, the best chip companies win on performance. Make Groq the default for fast inference.
Notable Investors: Disruptive, Tiger Global, D1 Capital Partners, T. Rowe Price, GIC
Valuation: $0.5B | Funding: $0.058B | Location: San Francisco, CA, USA
Team Size: ~37 employees | Category: AI Infrastructure
Product: Cloud platform for running open-source ML models via API - easy model deployment
Key Competitors: Modal, Baseten, Together AI, Fireworks AI, Hugging Face
Bull Case: Acquired by Cloudflare for $1B+; $58M raised; 50K+ models; easy deployment; a16z backing; made model deployment accessible
What Paul Graham Would Advise: You made model deployment easy - that's huge. Do things that don't scale: personally deploy models for your first users. Remember, the best developer tools remove friction. Make Replicate feel like magic.
Notable Investors: Andreessen Horowitz, Y Combinator, Sequoia Capital, Abstract Ventures
Valuation: $1.1B | Funding: $0.11B | Location: New York, NY, USA
Team Size: ~14 employees | Category: AI Infrastructure
Product: Serverless GPU compute for AI workloads - run AI code without managing infrastructure
Key Competitors: Replicate, Baseten, AWS Lambda, Google Cloud Run, Together AI
Bull Case: $1.1B valuation; $110M raised; 8-figure revenue; 14-person team; sub-second cold starts; Erik Bernhardsson (ex-Spotify); capital efficient
What Paul Graham Would Advise: 14 people to $1B+ valuation - that's the dream. Do things that don't scale: personally help users deploy their first workloads. Remember, the best infrastructure is invisible. Make Modal feel like having infinite GPUs on demand.
Notable Investors: Lux Capital, Redpoint Ventures, Amplify Partners, Addition
Valuation: $4.0B | Funding: $0.33B | Location: Redwood City, CA, USA
Team Size: ~148 employees | Category: AI Infrastructure
Product: Generative AI platform for fast model inference and deployment - compound AI systems
Key Competitors: OpenAI, Together AI, Replicate, Baseten, Modal
Bull Case: $4B valuation; $327M raised; 10T tokens processed daily; founded by Meta PyTorch team; fast inference; enterprise focus
What Paul Graham Would Advise: 10T tokens daily - you're handling serious scale. Do things that don't scale: personally optimize inference for your first enterprise customers. Remember, the best infrastructure companies win on reliability. Be the platform AI apps trust.
Notable Investors: Sequoia Capital, Benchmark, Lightspeed Venture Partners, Index Ventures
Valuation: $5.0B | Funding: $0.3B | Location: San Francisco, CA, USA
Team Size: ~147 employees | Category: AI Infrastructure
Product: ML infrastructure for model deployment and scaling - production ML platform
Key Competitors: Replicate, Modal, Seldon, Algorithmia, AWS SageMaker
Bull Case: $5B valuation; $300M+ raised; Series E led by IVP and CapitalG; enterprise focus; model deployment at scale; Greylock backing
What Paul Graham Would Advise: Production ML is hard - you're making it easier. Do things that don't scale: personally deploy models for your first enterprise customers. Remember, the best ML infrastructure is reliable. Be the platform enterprises trust for production.
Notable Investors: Greylock Partners, IVP, CapitalG, Conviction, Spark Capital
Valuation: $0.7B | Funding: $0.13B | Location: New York, NY, USA
Team Size: ~103 employees | Category: AI Legal
Product: AI agents for knowledge work in finance, law, and pharma - document analysis at scale
Key Competitors: Harvey, Primer, Eigen Technologies, Vectara, Kira Systems
Bull Case: $700M valuation; $130M+ raised; 33% of top asset managers as customers; 2,046% growth rate; $13M ARR; a16z backing
What Paul Graham Would Advise: Knowledge work is the perfect AI target - tons of documents, high stakes. Do things that don't scale: personally analyze documents for your first customers. Remember, the best AI finds insights humans miss. Make Hebbia feel like a research team on demand.
Notable Investors: Andreessen Horowitz, Index Ventures, Google Ventures, Peter Thiel
Valuation: $0.1B | Funding: $0.018B | Location: San Francisco, CA, USA
Team Size: ~3 employees | Category: AI Infrastructure
Product: Open-source embedding database for AI applications - simple vector storage
Key Competitors: Pinecone, Weaviate, Qdrant, LanceDB, Zilliz
Bull Case: $100M+ valuation; $18M raised; 3-person team; developer-friendly; simple API; adding memory to AI apps; YC backing
What Paul Graham Would Advise: 3 people, $18M raised - that's focus. Do things that don't scale: personally help every developer using Chroma. Remember, the best developer tools are simple. Make Chroma the easiest way to add memory to AI apps.
Notable Investors: Y Combinator, Quiet Capital, Bloomberg Beta, Naval Ravikant
Valuation: $0.2B | Funding: $0.027B | Location: San Francisco, CA, USA
Team Size: ~20 employees | Category: AI Developer Tools
Product: Data framework for LLM applications - connect external data to LLMs for RAG
Key Competitors: LangChain, Unstructured, Vercel AI SDK, FlowiseAI
Bull Case: $200M+ valuation; $27.5M raised; 20-person team; essential for RAG; LlamaCloud enterprise; data ingestion leader; Greylock backing
What Paul Graham Would Advise: Data + LLMs is the key to useful AI - you're solving it. Do things that don't scale: personally help developers build their first RAG apps. Remember, the best frameworks enable others to build. Be the data layer for the AI app ecosystem.
Notable Investors: Greylock Partners, Gradient Ventures, 01 Advisors, AIX Ventures
Valuation: $0.5B | Funding: $0.066B | Location: Los Angeles, CA, USA
Team Size: ~100 employees | Category: AI Video
Product: AI video generation platform with realistic AI avatars for marketing and training
Key Competitors: Synthesia, D-ID, Colossyan, Hour One, Elai
Bull Case: $500M+ valuation; $65.6M raised; Benchmark & Conviction co-led; realistic avatars; 100K+ customers; video at scale
What Paul Graham Would Advise: AI avatars are the future of video - you're making them accessible. Do things that don't scale: personally create videos for your first customers. Remember, the best AI video feels real. Make HeyGen avatars indistinguishable from humans.
Notable Investors: Benchmark, Conviction, SV Angel, Liquid 2 Ventures
Valuation: $1.5B | Funding: $0.47B | Location: San Francisco, CA, USA
Team Size: ~80 employees | Category: AI Coding
Product: AI coding startup with ultra-long context window models for code understanding
Key Competitors: GitHub Copilot, Cursor, Cognition AI, Poolside, Codeium
Bull Case: $1.5B+ valuation; $465M+ raised; ultra-long context windows; Eric Schmidt backing; CapitalG led; code understanding at scale
What Paul Graham Would Advise: Long context is the key to understanding code - you're pushing boundaries. Do things that don't scale: personally pair program with early users. Remember, the best coding AI understands entire codebases. Make Magic feel like a genius teammate.
Notable Investors: CapitalG, Eric Schmidt, Atlassian, Elad Gil, Nat Friedman
| Category | Companies | Total Valuation | % of Total |
|---|---|---|---|
| Foundation Model | 16 | $999.4B | 61.8% |
| AI Infrastructure/Data | 1 | $134.0B | 8.3% |
| AI Infrastructure | 18 | $62.0B | 3.8% |
| Autonomous Vehicles | 2 | $51.0B | 3.2% |
| AI Robotics | 3 | $46.5B | 2.9% |
| AI Coding | 7 | $56.3B | 3.5% |
| AI Security | 3 | $18.1B | 1.1% |
| AI Healthcare | 4 | $13.8B | 0.9% |
| AI Drug Discovery | 8 | $12.9B | 0.8% |
| AI Defense | 2 | $36.3B | 2.2% |
| AI Training Data | 1 | $29.0B | 1.8% |
| AI Search | 3 | $28.7B | 1.8% |
| AI Agents | 1 | $10.0B | 0.6% |
| AI Recruiting | 1 | $10.0B | 0.6% |
| AI Legal | 2 | $8.7B | 0.5% |
| AI Sales | 1 | $7.2B | 0.4% |
| AI Training | 1 | $6.3B | 0.4% |
| AI ML Platform | 1 | $6.3B | 0.4% |
| AI Media | 2 | $15.0B | 0.9% |
| AI Platform | 1 | $4.5B | 0.3% |
| AI Backend | 1 | $5.0B | 0.3% |
| AI Audio | 2 | $5.8B | 0.4% |
| AI DevTools | 2 | $6.4B | 0.4% |
| AI Video | 3 | $11.0B | 0.7% |
| AI Video/3D | 1 | $4.0B | 0.2% |
| AI Productivity | 2 | $13.0B | 0.8% |
| AI Bio | 1 | $10.0B | 0.6% |
| AI Translation | 1 | $2.0B | 0.1% |
| AI Support | 1 | $2.1B | 0.1% |
| AI Talent | 1 | $2.2B | 0.1% |
| AI Automation | 1 | $2.2B | 0.1% |
| AI Research | 1 | $3.1B | 0.2% |
| AI HR | 1 | $2.0B | 0.1% |
| AI Education | 1 | $1.0B | 0.1% |
| AI Data | 1 | $1.0B | 0.1% |
| AI MLOps | 1 | $1.0B | 0.1% |
| AI Development | 1 | $1.8B | 0.1% |
| AI Writing | 2 | $2.5B | 0.2% |
| Country | Companies | Total Valuation | % of Total |
|---|---|---|---|
| USA | 84 | $1,564.8B | 96.8% |
| France | 2 | $17.0B | 1.1% |
| China | 4 | $11.7B | 0.7% |
| UK | 3 | $7.5B | 0.5% |
| Canada | 1 | $5.5B | 0.3% |
| Germany | 1 | $2.0B | 0.1% |
| Japan | 1 | $3.0B | 0.2% |
| Israel | 1 | $1.2B | 0.1% |
| South Korea | 1 | $1.4B | 0.1% |
| Sweden | 1 | $1.8B | 0.1% |
| Rank | Company | Valuation | Category |
|---|---|---|---|
| 1 | OpenAI | $500.0B | Foundation Model |
| 2 | xAI | $230.0B | Foundation Model |
| 3 | Anthropic | $183.0B | Foundation Model |
| 4 | Databricks | $134.0B | AI Infrastructure/Data |
| 5 | Waymo | $45.0B | Autonomous Vehicles |
| 6 | Figure | $39.0B | AI Robotics |
| 7 | Safe Superintelligence (SSI) | $32.0B | Foundation Model |
| 8 | Anduril Industries | $31.0B | AI Defense |
| 9 | Anysphere (Cursor) | $29.3B | AI Coding |
| 10 | Scale AI | $29.0B | AI Training Data |
| 11 | Perplexity | $20.0B | AI Search |
| 12 | CoreWeave | $19.0B | AI Infrastructure |
| 13 | Mistral AI | $14.0B | Foundation Model |
| 14 | Thinking Machines Lab | $12.0B | Foundation Model |
| 15 | Notion | $11.0B | AI Productivity |
| 16 | Colossal Biosciences | $10.0B | AI Bio |
| 17 | Sierra | $10.0B | AI Agents |
| 18 | Mercor | $10.0B | AI Recruiting |
| 19 | Cognition AI | $9.8B | AI Coding |
| 20 | Cyera | $9.0B | AI Security |
Foundation model companies represent 16 of the top 100 but account for $999.4B in valuation (61.8% of total). The winner-take-most dynamics in AI are stark.
84% of companies are US-based, representing 96.8% of total valuation. The AI revolution is largely an American phenomenon, though Europe (Mistral, Poolside) and China (Moonshot, MiniMax) are emerging.
AI infrastructure companies (compute, databases, deployment) represent 18 companies with $62B in valuation. As the saying goes, in a gold rush, sell picks and shovels.
Legal (Harvey), healthcare (Abridge), and defense (Anduril, Shield AI) are seeing massive AI-native disruption. Vertical-specific AI is proving more valuable than general-purpose tools.
- Most efficient: Midjourney ($500M revenue, $0 funding)
- Highest funding: OpenAI ($59B raised, $500B valuation)
- Best ratio: Cursor ($29B valuation on $2.3B funding = 12.6x)
AI coding tools (Cursor, Cognition, Poolside, Codeium, Magic) represent $56B+ in valuation. The way software is built is being fundamentally rewritten.
8 companies in AI drug discovery with $12.9B in valuation. Insitro, Isomorphic Labs, and Recursion are betting AI can revolutionize pharmaceutical R&D.
Anduril ($31B) and Shield AI ($5.3B) prove that defense technology is a viable (and lucrative) AI application. Software is eating defense.
-
"Do Things That Don't Scale" (100% of companies)
- Personal onboarding for first customers
- Manual work to understand problems deeply
- Hand-holding early adopters
-
"Solve Your Own Problem" (Foundation models, coding tools)
- OpenAI built ChatGPT because they wanted better AI
- Cursor built what they wished they had as developers
- Harvey built what lawyers needed
-
"The Schlep is the Moat" (Infrastructure, security, healthcare)
- Data labeling (Scale AI) is tedious but essential
- Compliance (Vanta) is boring but necessary
- Healthcare AI requires regulatory navigation
-
"Live in the Future" (Robotics, autonomous vehicles, drug discovery)
- Figure is building the future of physical labor
- Waymo is living in a world with robotaxis
- Insitro is creating the future of drug discovery
-
"Relentlessly Resourceful" (All companies)
- Every startup faces impossible odds
- The winners find ways to make things happen
- Funding helps, but execution matters more
Companies were included if they meet ALL of the following:
- AI is core to the business model (not a feature)
- Building proprietary AI technology (not just using APIs)
- Recent funding (2023-2025) or significant traction
- Genuine innovation in AI (not marketing fluff)
- CB Insights AI 100 (2024, 2025)
- Crunchbase Unicorn Board (January 2025)
- PitchBook Data (2025)
- TechCrunch Funding Reports (2023-2025)
- Company Press Releases (2024-2025)
- Exploding Topics - Generative AI Startups (2025)
- Latent Space Podcast - AI Engineer Community
- VC Portfolio Pages - Sequoia, Benchmark, Greylock, Conviction
- Valuations represent post-money valuations from latest funding rounds
- Figures are the most recent publicly available as of January 2025
- Some companies may have raised additional rounds not yet reflected
- Chinese companies may have additional undisclosed funding
The following were EXCLUDED as AI pretenders:
- C3.ai (enterprise software with AI marketing)
- IBM Watson (legacy tech rebranded)
- Salesforce Einstein (feature, not product)
- Oracle AI (database company)
- SAP AI (ERP company)
- Any company where AI is <50% of value proposition
This report was compiled by synthesizing data from multiple authoritative sources to create the definitive list of genuine AI-native startups. The goal is to separate the signal from the noise in the AI hype cycle.
Key Principles:
- Accuracy over hype: Real funding figures, not marketing claims
- Quality over quantity: 100 genuine companies > 500 pretenders
- Actionable insights: Paul Graham advice tailored to each company
- Transparency: Clear methodology and sources
Report compiled: January 2025
Data sources: CB Insights, Crunchbase, PitchBook, TechCrunch, Company Press Releases
For updates and corrections: Please refer to the latest funding announcements
END OF REPORT