14 categories and 78 products and companies, charted by an ai agent.
Frontier language models are large neural networks trained on broad text data that perform language understanding and generation tasks across multiple domains. Researchers, product teams, and enterprises use them to power conversational interfaces, content generation, reasoning tasks, and custom applications.
claude opus 4.8 · claude sonnet 4 · claude haiku 3.5 · gpt-4o · gemini 2.5 pro · mistral large · llama 3.3 · deepseek v3
Tools that enable Claude (Anthropic's AI model) to write, test, execute, and iterate code autonomously or with human direction. Software engineers, data scientists, and developers use these to automate coding tasks, debugging, and software development workflows.
claude code · cursor · github copilot · windsurf · aider · devin · cline · replit · openai codex
anthropic api · amazon bedrock · google cloud vertex ai · microsoft foundry · together ai · fireworks ai
AI chat assistants are conversational interfaces that accept text input and generate text responses, used by consumers and enterprises for question-answering, writing, analysis, and coding tasks. Users interact through web or mobile apps to have multi-turn conversations with language models.
claude.ai · chatgpt · gemini · copilot · perplexity · meta ai · gemini app · microsoft copilot · le chat
AI agent frameworks are software platforms and libraries that enable developers to build autonomous agents that perceive environments, make decisions, and take actions toward goals. Data scientists, ML engineers, and application developers use them to create multi-step workflows, tool-use systems, and reasoning loops.
langchain · langgraph · autogen · crewai · model context protocol · llamaindex · vercel ai sdk
Enterprise AI productivity tools help knowledge workers automate writing, analysis, coding, and research tasks within business workflows. Companies use these to augment employee output across sales, marketing, engineering, and operations teams.
claude for work (team & enterprise) · microsoft 365 copilot · notion ai · salesforce einstein · glean · writer
claude research mode · openai deep research · perplexity deep research · gemini deep research · you.com research
AI education tools are software products that teach programming, data science, machine learning, or other technical skills using AI tutors, automated grading, or personalized learning paths. Students, bootcamp participants, and corporate training teams use them to learn coding and AI concepts.
claude for education · khanmigo · chegg mathway ai · synthesis tutor
anthropic alignment science · openai safety systems · arc-agi evals · deepmind safety team
Software that takes autonomous actions on a user's computer—executing tasks, navigating interfaces, running commands—based on natural language instructions. Used by enterprise teams, software developers, and knowledge workers to automate repetitive digital work.
claude cowork
Foundation models are large neural networks trained on broad data that can be adapted to many downstream tasks. Developers, researchers, and enterprises use them via APIs or by running them locally to power applications from chatbots to code generation to reasoning tasks.
anthropic claude · openai · google gemini · meta llama · mistral · deepseek · xai grok
Cloud model platforms provide hosted infrastructure and APIs for training, fine-tuning, and deploying large language models without managing hardware directly. Data scientists, ML engineers, and enterprises use these to build applications with LLMs.
google vertex ai · azure ai foundry · databricks · snowflake cortex
Legal AI tools help lawyers and legal teams draft documents, review contracts, conduct legal research, and manage cases. In-house counsel, law firms, and legal departments use these systems to automate document assembly and contract analysis.
harvey · legora · casetext · evenup
AI safety evaluation covers tools and services that test, audit, and measure risks in AI systems including adversarial robustness, alignment, bias, and harmful outputs. ML teams, safety researchers, and enterprises use these to assess model behavior before deployment.
metr · apollo research · scale ai · humanloop