The AI Tools Landscape
← UP Academy · Free course · 5 lessons
The AI Tools Landscape
Snapshot as of Oct 2026. Products change fast, so check vendor sites. UP has no paid relationship with these vendors unless disclosed.
Lesson 1: Major model families
- Closed / commercial: OpenAI GPT (ChatGPT), Anthropic Claude, Google Gemini, xAI Grok.
- Open-weight: Meta Llama, Mistral, DeepSeek, Alibaba Qwen, Google Gemma. You can download these and run them yourself, subject to each licence.
- How to choose: quality on your tasks, price, privacy, context length, and tool support. Test with your own examples.
Lesson 2: Agent frameworks (for builders)
LangChain / LangGraph, LlamaIndex, CrewAI, Microsoft AutoGen / Agent Framework, Hugging Face smolagents, and the OpenAI Agents SDK. MCP (Model Context Protocol) is the emerging standard for connecting tools to models. Anthropic’s advice holds: start simple, because many tasks need a workflow, not an agent.
Lesson 3: No-code and low-code builders
Zapier, Make and n8n (n8n can be self-hosted) connect apps with AI steps. Custom assistants (GPTs, Gemini Gems, Claude Projects) need no coding at all. These are good for small-business automations and side services.
Lesson 4: Open-source tooling
Hugging Face (models, datasets, the free Agents Course), Ollama and LM Studio (run models locally), and vector databases for search over your own documents.
Lesson 5: Safety checklist for any tool
Read the data policy, keep secrets out, give agents the fewest permissions possible, require human approval for anything irreversible, and watch for prompt injection (OWASP LLM01).
Not accredited. UP Academy courses and certificates are educational and are not accredited by any school, licensing body or government agency. Nothing here is legal, financial or medical advice.
AI-agent disclosure. UP Academy content is researched and drafted with help from AI agents and reviewed by Undeniable Progress. AI-generated videos are labelled as such.
