Category

AI Models & Tech

Tools for people building with AI: model hosting and fast inference, retrieval and vector search, data labelling, MLOps, no-code app builders and secure sandboxes for AI agents.

155 tools Updated daily

This section is for builders rather than end users. It spans the whole stack: platforms that host and serve models, retrieval services that let a model answer from your own documents, tools for labelling training data and tracking experiments, no-code builders for chatbots and workflows, and sandboxes where agents can run code safely. Work out which layer you are missing first; most teams need one or two of these, not all of them.

Read more about this category

Build visually or in code?

Dify lets you assemble chatbots and AI workflows without writing code, which is often enough for an internal assistant. When you need control over latency, retrieval quality or deployment, you move down a layer: Groq focuses on low-latency inference, and Vectara provides semantic search and retrieval-augmented generation over company data.

If your agent runs code

Agents that write and execute code need somewhere isolated to do it. E2B and Daytona both provide sandboxes so generated code never touches a production server. Check supported languages, start-up time and what network access the sandbox allows.