Reference
A categorized reference of AI companies, from frontier model labs to applied AI software. Built for executives evaluating where to invest, partner, or build.
One outer rectangle, 14 color-coded category rectangles sized by combined vendor value. Within each category, every vendor logo is sized proportionally using log-scaled company-wide market cap (public) or last private valuation (private), May 2026 estimates. A few vendors appear in multiple categories when their AI product spans them.
No vendors match your search. Try a different term, or submit one we should add.
The handful of organizations training the largest, most capable general-purpose models. These labs anchor the cost curve, the safety frontier, and most enterprise AI roadmaps.
Maker of the GPT and o-series models, ChatGPT, Sora, and the OpenAI API. Anchored by a deep partnership with Microsoft.
Creator of Claude, with a research focus on AI safety, interpretability, and building steerable models for high-stakes enterprise work.
Google's combined AI research division behind Gemini, AlphaFold, and Veo. Deeply integrated into Search, Workspace, and Google Cloud.
Meta's research arm and home of the Llama family of open-weight models. The reference open-source ecosystem for enterprise self-hosting.
Elon Musk's AI lab behind the Grok model family, integrated across X (formerly Twitter) and exposed via its own API.
Paris-based lab producing both open-weight and frontier-tier models with a strong focus on European data sovereignty.
Enterprise-focused foundation models (Command, Embed, Rerank) optimized for retrieval-augmented generation in regulated industries.
Israeli lab behind the Jamba family of hybrid Mamba/Transformer models, focused on long-context enterprise use cases.
Multimodal foundation model lab (Reka Core, Flash, Edge) targeting on-prem and edge deployment for enterprise.
Pivoted to enterprise after its commercial team joined Microsoft. Now offers Inflection-tuned models for regulated workloads.
Apple's on-device foundation models powering Siri, writing tools, and image generation across iPhone, iPad, and Mac.
ByteDance's flagship model family powering Doubao chatbot, dominant in China with multimodal and long-context capabilities.
Consumer-facing AI chatbots backed by their own foundation models. Multi-billion-dollar Google partnership.
Open-weight, niche, or non-Western frontier models. Often selected for cost, control, sovereignty, or specialized capability.
The default hub for open-source model weights, datasets, and training tooling. Effectively the GitHub of machine learning.
Data and AI platform with the DBRX open model and Mosaic AI training stack. Strong story for fine-tuning on proprietary data.
Open-source generative media (Stable Diffusion, Stable Audio, Stable Video). The reference open image-generation lineage.
Founded by ex-Stability researchers; the FLUX family is a current state-of-the-art open image-generation lineage.
German foundation model company focused on sovereign AI deployments for European government and regulated enterprise.
Chinese lab whose efficient reasoning models reset assumptions on training cost and open-weight performance.
Alibaba's open-weight Qwen model family, a major non-Western foundation model series with strong multilingual coverage.
Maker of the Kimi long-context model family. Notable for strong long-document and reasoning benchmarks.
Founded by Kai-Fu Lee. Maker of the Yi family of open-weight bilingual models targeting English and Chinese workloads.
Action-oriented multimodal models for browser and desktop automation. Now operating largely as part of Amazon.
Local-first runtime for open-weight models. The simplest path to running Llama, Mistral, and Qwen on your laptop.
Desktop application for discovering, downloading, and chatting with local open-source LLMs across Windows, macOS, and Linux.
Where most enterprises actually consume AI: hyperscalers and cloud platforms that bundle compute, model gardens, training infrastructure, and managed services.
Amazon's managed model garden (Anthropic, Meta, Mistral, Amazon Nova, Titan) plus the SageMaker training and serving stack.
Azure OpenAI Service, Azure AI Foundry, and Copilot Studio. The default enterprise on-ramp for OpenAI models.
Google Cloud's AI platform: Gemini, Imagen, Veo, Model Garden, and managed training, all wired into BigQuery and TPU v5/Trillium.
IBM's enterprise AI platform built around the Granite model family, governance tooling, and hybrid-cloud deployment.
OCI Generative AI Service plus Oracle's high-density GPU clusters underpinning much of its AI infrastructure backlog.
The leading non-US hyperscaler in many APAC markets, with deep integration into the Qwen model family.
Tencent's enterprise AI platform with Hunyuan models, vector DB, and agent tooling for Asia-Pacific markets.
Serverless inference platform running open-source models on Cloudflare's global edge network.
The supply side of AI: silicon, GPU clouds, and inference platforms. The economics here flow downstream into every model and product.
The dominant AI accelerator vendor. CUDA, H100/H200/Blackwell, NIM microservices, and the broader Omniverse / DGX stack.
The Instinct MI300/MI350/MI400 GPU line plus ROCm software. The credible second source named in most major AI build-outs.
Intel's Gaudi accelerator line plus Xeon AI extensions. A challenger position with selective enterprise traction.
Wafer-scale AI engines targeting the largest single-system training and ultra-fast inference workloads.
Custom LPU inference silicon delivering very high tokens-per-second economics for production LLM serving.
Reconfigurable dataflow accelerators packaged as a full enterprise AI stack, often deployed on-prem.
Open-architecture AI silicon (Wormhole, Blackhole) with a RISC-V CPU strategy led by Jim Keller.
UK-based IPU silicon vendor, now part of SoftBank. Refocused on next-generation AI compute architectures.
Pure-play GPU hyperscaler. The reference example of NVIDIA-aligned, AI-native cloud capacity at scale.
GPU cloud and on-prem AI infrastructure. Popular with research labs and mid-market AI-native companies.
Vertically integrated AI cloud building dedicated power and data centers, including stranded-energy and gas-to-AI sites.
Inference and fine-tuning cloud for open-source models. Often the cheapest production path for Llama-class workloads.
High-performance inference platform with strong throughput on open-weight LLMs and multimodal models.
Serverless GPU compute with a Python-first developer experience. Popular for batch inference and fine-tuning jobs.
On-demand and spot GPU rental targeted at developers, hobbyists, and AI-native startups.
Commercial Ray distribution. The reference framework for distributed training, RL, and large-scale inference workloads.
Transformer-specific ASIC (Sohu) burning the architecture into silicon for an order-of-magnitude inference speedup.
The plumbing layer underneath production AI: vector databases, retrieval, observability, evaluation, and the orchestration stack.
Managed vector database widely used for RAG. The reference enterprise serverless vector store.
Open-source vector database with strong hybrid search and graph features for retrieval-augmented applications.
High-performance open-source vector database written in Rust, popular for self-hosted and on-prem RAG.
Developer-friendly embeddings database; the default choice for prototype RAG pipelines and notebooks.
Commercial offering and steward of Milvus, the large-scale open-source vector database used heavily in APAC.
Embedded vector database built on the Lance columnar format. Often used inside multimodal data pipelines.
Open-source orchestration framework plus a hosted observability and evaluation product (LangSmith) for LLM applications.
Data framework specialized in RAG and agentic document workflows. The other dominant open-source orchestration choice.
The dominant ML experiment-tracking and observability platform. Widely used inside frontier labs and enterprise ML teams.
End-to-end ML and LLM experiment management with a strong story around model registry and production monitoring.
Production ML and LLM observability, with the open-source Phoenix tooling and a hosted enterprise monitoring product.
Model observability and explainability for regulated industries. Strong story around bias, drift, and audit.
LLM evaluation, hallucination detection, and guardrails specifically designed for production GenAI applications.
LLM evaluation and prompt management platform. Targets product teams shipping LLM features at scale.
Snowflake's LLM and ML feature set, bringing model serving to the data warehouse rather than moving data to the model.
Data labeling, evaluation, and RLHF infrastructure. A core supplier to frontier labs and the US public sector.
Managed data movement and AI-ready pipelines feeding warehouses and lakehouses for retrieval and analytics.
Vector search and AI-native retrieval inside the Elasticsearch platform, including ELSER semantic ranking.
Where AI lands inside the enterprise: horizontal SaaS platforms layering AI on top of CRM, ERP, HCM, ITSM, and analytics. Mostly incumbents under pressure to ship.
Foundry plus the AI Platform (AIP). Strong enterprise traction in industrial, defense, healthcare, and intelligence work.
Salesforce's agent platform, evolution of Einstein. Tightly coupled to the Data Cloud and CRM workflow surface.
AI features layered over ITSM, HR, and customer workflows. Notable for its workflow-anchored, on-platform value capture.
Embedded AI across HCM and Finance. Workday Illuminate frames their generative AI strategy on top of customer data.
SAP's AI copilot and generative AI hub embedded across S/4HANA, Ariba, and SuccessFactors.
Application-layer enterprise AI focused on industrial, energy, defense, and government use cases.
End-to-end ML and GenAI platform for governed enterprise model development, deployment, and monitoring.
AutoML, generative AI, and document intelligence platform with strong adoption in financial services.
Full-stack generative AI platform with its own Palmyra models, focused on regulated enterprise content workflows.
Cohere's secure enterprise AI workspace, deployable in customer cloud or on-prem for regulated and sovereign workloads.
Enterprise AI search and assistant across Slack, Drive, Confluence, Salesforce, and the rest of the SaaS surface.
Enterprise AI copilot for IT, HR, and employee support. Acquired by ServiceNow in 2025.
Enterprise AI platform connecting LLMs to internal knowledge, with agents and workflows tailored per team.
Conversational AI and agent platform for IT, HR, and customer service automation in large enterprises.
AI-driven data security and management with Gaia conversational assistant over enterprise backup data.
Native AI summarization, Q&A, and metadata extraction across documents stored in Box.
AI coding assistants and autonomous engineering agents. The most validated horizontal AI use case in the enterprise today.
The original AI coding assistant. Now multi-model, with Copilot Workspace and Spark agents extending it beyond inline completion.
AI-native fork of VS Code that became the default IDE for many AI-forward engineering teams.
Anthropic's agentic coding tool. Runs in the terminal, IDE, or CI to drive multi-step engineering tasks.
Agentic AI IDE built by Codeium, with a strong enterprise self-hosting story.
Privacy- and compliance-focused AI coding assistant, with a self-hosted enterprise story.
Code intelligence platform with the Cody assistant and Amp agent, focused on large-scale enterprise codebases.
Browser-native development environment with an autonomous agent that can stand up full applications from a prompt.
Maker of Devin, a fully autonomous software engineering agent positioned for end-to-end ticket execution.
Frontier coding model lab focused on extreme long-context reasoning over very large codebases.
Frontier code-specialist lab building proprietary models and an enterprise platform for software development.
Enterprise-grade AI coding platform that indexes the full codebase to provide context-aware completions and chat.
Prompt-to-app builder for full-stack web applications. A leading no-code GenAI development surface.
Vercel's AI app builder for React and Next.js. The reference design-to-code GenAI tool for frontend teams.
Browser-based AI development environment that ships full-stack apps without a local toolchain.
Browser-based AI app builder for shipping full-stack web applications from natural-language prompts. Aimed at fast prototyping and non-developer founders.
Open-source coding assistant for VS Code and JetBrains. Customizable with any LLM and integrates with internal codebases.
No-code AI app builder. Acquired by Wix in 2025 to power vibe-coded SaaS.
Frameworks and platforms for building autonomous or semi-autonomous AI agents that execute multi-step business workflows.
Open-source multi-agent framework plus a commercial enterprise platform. Popular for structured agentic workflows.
Build-your-own AI assistant platform. Strong fit for SMB and ops teams automating repetitive knowledge work.
Zapier's no-code agent builder, layered on top of its 7,000+ app integration library.
Browser-based AI automation platform for repetitive web workflows. Strong fit for sales, recruiting, and ops teams.
Open-source workflow automation platform with first-class AI and agent nodes. Often self-hosted by enterprise IT.
RPA leader extending into agentic automation, combining classic process automation with LLM-driven decisioning.
Web automation agent platform that operates real browser sessions on the user's behalf.
No-code AI workforce builder for sales, marketing, and operations. Strong traction with revenue teams.
LangChain's stateful, graph-based framework for building reliable, controllable agent workflows.
Headless browser infrastructure purpose-built for AI agents. Underpins many production browsing agents.
OS-layer agent startup founded by senior alumni from Google, OpenAI, and Meta. Targeting an agent-native operating environment.
Conversational AI platform for customer experience, founded by Bret Taylor. Production agents at Sonos, ADT, and others.
AI agents for enterprise customer support. Deployed by Bilt, Eventbrite, and Substack to handle complex multi-turn cases.
AI-powered customer service automation platform. Used by Square, Verizon, and Meta to resolve tickets without human handoff.
Visual automation platform with AI agents for cross-app workflows. Owned by Celonis.
No-code AI workflow builder for repeatable knowledge-work automations.
Enterprise iPaaS that has pivoted to AI agents for integrating and automating across SaaS.
Marketplace of AI agents plus a no-code builder, from HubSpot CTO Dharmesh Shah.
AI-native search, research copilots, and document intelligence. The most direct competitive challenge to traditional web search.
Answer engine combining real-time web retrieval with cited LLM synthesis. The reference AI search consumer product.
AI search with consumer and enterprise products, plus an API for retrieval-grounded generation.
Neural search API designed for AI agents and research workflows. Embeddings-first rather than keyword-first.
AI search engine over peer-reviewed scientific literature. Used heavily in healthcare, life sciences, and academic research.
Document intelligence and research copilot for finance, law, and consulting. Strong on reading large document sets.
RAG-as-a-service platform with built-in retrieval, generation, and hallucination detection.
AI research assistant for systematic literature review. Used widely in academia and evidence-based policy work.
AI assistant for academic and technical paper reading, summarization, and citation tracking.
AI search engine optimized for developers and technical questions. Returns code, citations, and explanations side by side.
Privacy-focused AI assistant built into the Brave browser, offering chat, summarization, and code help without account or tracking.
OpenAI's AI-native browser with built-in agentic browsing and ChatGPT integration.
Generative media tools across image, video, voice, and music. Increasingly the front line for marketing, content, and creative production.
The aesthetic benchmark for AI image generation. Closed model, distinctive style, strong creative-pro adoption.
AI image generation with a strong reputation for legible in-image text. Popular for marketing and design surfaces.
Adobe's commercially safe generative model family, embedded across Photoshop, Illustrator, Premiere, and Express.
Generative design tool focused on brand-consistent vector and raster output for product and marketing teams.
Real-time generative image and video tool with strong creative-pro and design-studio adoption.
Generative video pioneer (Gen-3, Gen-4). Deep integrations into film, advertising, and broadcast pipelines.
Consumer and prosumer generative video tool with strong viral momentum on social platforms.
Maker of the Dream Machine generative video model and Genie 3D capture tools.
AI video generation platform built around realistic AI avatars. Heavy enterprise traction in L&D and corporate comms.
AI avatar and translation video platform popular with sales, marketing, and personalized-video use cases.
The reference AI voice synthesis platform. Used across audiobooks, dubbing, gaming, and conversational agents.
Generative music creation platform producing full songs, including vocals, from text prompts.
Generative music platform competing with Suno on audio quality and prompt expressiveness.
AI-first audio and video editor with transcription, voice cloning, and overdub. Popular with podcasters and creators.
Speech AI API for transcription, diarization, and speech understanding. Popular with developer and contact-center workloads.
Real-time speech-to-text and voice AI APIs optimized for low-latency conversational use cases.
Design platform with Magic Studio AI suite for image generation, copy, video, and brand-consistent visuals. Acquired Leonardo AI in 2024.
Image generation platform popular for game art, design assets, and consistent character workflows. Owned by Canva since 2024.
AI-native video creation app: avatars, dubbing, captions, and short-form video editing tuned for creators and marketers.
Enterprise generative AI for on-brand content creation at scale across formats.
AI-powered video editing platform with avatars, dubbing, and translation.
AI ad creative production and optimization platform for performance marketing.
AI features and standalone tools that reshape how individual knowledge workers write, meet, and organize. The fastest-adopted layer of GenAI in most enterprises.
The Microsoft 365 Copilot suite across Outlook, Word, Excel, PowerPoint, Teams, and Copilot Studio.
Gemini features in Gmail, Docs, Sheets, Slides, and Meet. Default AI surface for Google-stack enterprises.
AI features and search inside Notion's docs, projects, and databases. Increasingly used as a horizontal AI layer for SMB.
Atlassian's enterprise AI search and agent layer over Jira, Confluence, and connected SaaS tools.
Meeting transcription, summarization, and AI assistant. One of the longest-standing meeting AI products.
AI dictation tool that converts speech to polished, context-aware writing across any app. Strong founder and exec adoption.
Meeting recorder and AI notetaker with strong CRM and revenue-team integrations.
AI notepad for meetings that augments your own notes rather than replacing them. Strong founder and exec adoption.
Free AI meeting recorder, summarizer, and CRM logger. Heavy SMB and revenue-ops adoption.
Meeting analytics, summary, and engagement scoring across Zoom, Meet, Webex, and Teams.
AI-native personal knowledge base that auto-organizes notes, meetings, and personal context.
Networked notes app with embedded AI for thinking, journaling, and personal CRM.
AI presentation, doc, and webpage builder. The reference fast-iteration alternative to PowerPoint and Google Slides.
AI-native presentation tool, repositioned as a sales-focused account intelligence and outbound platform.
AI presentation builder that turns a short prompt into a complete deck with speaker notes. Popular SMB alternative to Gamma.
AI-augmented note and knowledge tool that turns unstructured notes into structured nodes, tasks, and projects.
Voice AI for noise cancellation, accent translation, and real-time meeting transcription.
AI post-production for podcasts and recordings, automatically removing fillers, mouth sounds, and silence.
Personal AI assistant with a searchable memory of what you have seen, said, and heard on your computer.
AI-native email client built for speed. Acquired by Grammarly in 2025.
Google's AI research notebook that grounds answers in user-provided source documents.
AI applied to revenue functions: copywriting, content production, prospecting, customer support, and conversation intelligence.
Marketing-focused generative AI platform for brand-aligned content, campaigns, and copy at scale.
GenAI workflow platform for go-to-market teams, with a strong agent and workflow story for revenue ops.
Performance-focused AI copy platform. Predictive scoring of marketing copy against custom audiences and brand voice.
HubSpot's AI brand: copilots, agents, and intelligence layered across the marketing, sales, and service hubs.
Revenue intelligence platform analyzing sales calls, emails, and pipeline data for forecasting and coaching.
AI-powered revenue platform for forecasting, pipeline, and rep productivity. Heavy enterprise traction.
AI-driven account-based marketing and sales intelligence. Strong story on intent data and predictive scoring.
AI customer support agent embedded in Intercom. Highly visible deployment in front-line CX.
AI agents, copilots, and intelligence layered across the Zendesk customer service suite.
AI agents and assist for support teams, with deep integrations into existing CX platforms.
Real-time agent assist and conversation intelligence for sales and contact-center teams.
Conversation intelligence and AI coaching for contact centers, with strong analytics and QA workflows.
AI sales agents for prospecting, sequencing, and personalized outbound at scale.
Autonomous digital workers (Alice, Jordan) for sales prospecting and SDR-style work.
AI control layer for paid search optimization. From MotiveMetrics.
All-in-one AI-powered GTM platform combining B2B data, prospecting, sequencing, and conversation intelligence.
AI-enriched data engine for GTM teams to research, segment, and personalize outbound at scale.
Sales engagement platform using AI for conversation insights, deal scoring, and rep coaching.
Sales execution platform applying AI to sequence optimization, forecasting, and rep guidance.
Generative AI for marketing language, producing message variants optimized to drive measurable response.
AI sales email coach that scores and rewrites emails in real time as reps draft them.
AI-personalized video at scale, generating one-to-many video campaigns with synthesized faces and voices.
End-to-end AI sales platform with autonomous Ava digital sales rep handling prospecting and outbound.
AI sales development rep that personalizes outbound at scale across email and LinkedIn.
AI-powered sales prospecting with intent signals, account research, and enriched B2B data.
AI SDR that researches target accounts and writes personalized outbound sequences.
Piper, an AI SDR that holds real conversations with website visitors and books qualified meetings.
Enterprise AI customer service platform with multi-step reasoning over knowledge and tools.
Conversational AI voice agents that resolve customer service calls end-to-end.
Enterprise conversational AI platform for customer and employee experience automation.
Deeply specialized AI built for a specific industry workflow. Often the most defensible category against horizontal LLMs alone.
Generative AI platform purpose-built for elite law firms and corporate legal teams. The reference legal AI win.
Generative AI contract drafting and review for transactional lawyers, embedded in Microsoft Word.
LexisNexis's generative AI legal research and drafting product, layered on top of its citation graph.
CoCounsel (originally Casetext) plus broader AI features across Westlaw, Practical Law, and tax workflows.
Precision medicine and clinical AI built on a large library of multimodal patient data.
AI-native drug discovery platform built around large-scale phenomic and molecular datasets.
Machine learning-driven drug discovery focused on human biology data and induced pluripotent stem cells.
Alphabet drug discovery spinout building on AlphaFold, with major partnerships with Eli Lilly and Novartis.
Generative AI healthcare staffing agency operating non-diagnostic, patient-facing AI agents.
AI clinical documentation that converts patient-clinician conversations into structured medical notes.
Generative AI scribe and operational platform for physicians and large health systems.
Voice-enabled AI clinical assistant for documentation, coding, and Q&A inside the EHR.
Talent intelligence platform applying AI to recruiting, talent management, and workforce planning.
AI-driven hiring assessments, video interviews, and conversational chat for high-volume recruiting.
Conversational AI assistant Olivia, used heavily for high-volume hourly recruiting in retail and food service.
S&P Global's AI research division, building NLP and ML tools for capital markets and financial documents.
AI-powered market intelligence search across filings, transcripts, expert calls, and news for finance and corporate strategy.
Foundation-model-era robotics: humanoids, manipulation, and embodied AI. The longest-cycle but highest-stakes frontier of AI deployment.
Humanoid robot company (Figure 02 / 03) with high-profile commercial pilots in logistics and manufacturing.
Maker of the Atlas humanoid and Spot quadruped. Now part of Hyundai's robotics push.
Norwegian-American humanoid company building Eve and Neo for home and commercial settings.
Maker of Digit, a humanoid robot with deployments in warehousing and logistics for customers including GXO.
Humanoid robotics company developing Apollo for industrial and logistics environments.
Canadian humanoid robotics company building general-purpose humanoid systems with the Phoenix platform.
Robotics foundation model company building a general-purpose brain across robot embodiments.
Foundation-model lab for robotics. Notable for its π0 model targeting general-purpose robot manipulation.
Industrial robotic manipulation foundation models. Founding team largely joined Amazon in 2024.
Know an AI app, tool, service, or website that should be in the landscape? Submit it for review and we'll evaluate it for inclusion in the next monthly refresh.
