AI Tool Category Taxonomy
A reference for how Hokai categorizes the AI tool landscape. Organized by architecture type, use case, and deployment. Each category has a definition and 2–3 example tools.
By Architecture Type
Foundation Models — Core LLMs and multi-modal models. The brains. Examples: ChatGPT, Claude, Gemini.
AI Agents — Tools that take autonomous actions. Use APIs, browse, execute. Examples: Cursor agent mode, Claude Code, workflow agents.
Workflow Platforms — Connect AI to other software. Triggers, actions, automation. Examples: Zapier, Make, n8n.
Developer Tools — AI for coding. IDEs, copilots, terminals. Examples: Cursor, GitHub Copilot, Replit.
Vertical SaaS — Industry-specific AI. CRM, HR, legal, healthcare. Examples: Salesforce Einstein, Workday, legal AI platforms.
By Use Case
Writing — Drafting, editing, copywriting. Examples: Jasper, Copy.ai, Notion AI.
Image Generation — Create images from prompts. Examples: Midjourney, DALL-E, Flux.
Video — Generate or edit video. Examples: Runway, Pika, Descript.
Audio — Voice, music, transcription. Examples: ElevenLabs, Suno, Otter.
Code — Coding assistance and generation. Examples: Cursor, Copilot, Codeium.
Data Analytics — Analysis, visualization, reporting. Examples: Tableau AI, Power BI, AI analytics tools.
Customer Support — Chatbots, triage, knowledge base. Examples: Intercom Fin, Zendesk AI, Freshdesk.
Marketing — Content, SEO, ads, email. Examples: Jasper, Surfer, HubSpot AI.
Sales — Prospecting, outreach, CRM. Examples: Outreach, Gong, Salesforce Einstein.
HR and Recruiting — Screening, scheduling, sourcing. Examples: Greenhouse, Lever, Gem.
Legal — Contract review, research. Examples: Kira, Luminance, Casetext.
Finance — Analysis, forecasting, extraction. Examples: Document AI, financial AI platforms.
Education — Learning, tutoring, assessment. Examples: Khanmigo, Duolingo Max.
Research — Literature review, synthesis. Examples: Elicit, Perplexity, Semantic Scholar.
Productivity — Tasks, notes, scheduling. Examples: Notion AI, Todoist, Motion.
Design — Layout, assets, prototyping. Examples: Figma AI, Canva, Uizard.
By Deployment
Cloud / SaaS — Hosted. Access via web or API. Most common. Examples: ChatGPT, Claude, most tools.
Self-Hosted — Run on your infrastructure. Data stays local. Examples: Llama, Mistral, n8n self-hosted.
API-Only — No UI. Integrate via API. Examples: OpenAI API, Anthropic API, embedding APIs.
Desktop — Install on your machine. Examples: Cursor, Claude desktop, some coding tools.
Mobile — Native mobile apps. Examples: ChatGPT app, many consumer AI apps.
Browser Extension — Works inside the browser. Examples: Grammarly, various AI assistants.
The Bottom Line
This taxonomy organizes AI tools by architecture, use case, and deployment. Use it to navigate the directory, filter tools, and understand the landscape. Hokai uses these categories for navigation and Smart Match.