Skip to content

10 curated AI tools for Agent Platforms

Agent platforms give developers the frameworks and self-hosted environment to build, orchestrate and run their own agents on their own terms.

More on this category

What these tools are

Agent platforms are the frameworks and self-hosted environments developers use to build agents: defining tools, wiring flows, connecting models and recording what happened. They are for teams that want control over models, data and deployment.

What they usually do

Orchestrate multiple models and tools into a flow, expose internal APIs to an agent, add retrieval over your own documents, manage sessions and memory, and deploy the whole thing on your own infrastructure.

Strengths and limits

The upside is control, auditability and access to private data. The cost is engineering: model choice, prompt maintenance, upgrades and monitoring all need an owner.

How pricing works

The frameworks are usually open source and free, so cost lands on model calls and servers. Hosted versions charge by executions, seats or credits; enterprise is quote-based.

How to choose
  • Check which models and vector stores it supports, and whether you can swap them
  • Confirm observability: traces, logs and cost tracking
  • Review permissions and multi-tenancy if a team will share it
  • Weigh the operational burden of self-hosting

Maintained by steven · About our listings →