🚀 EDDI 6.5.0

19 LLM providers, knowledge bases that crawl your site and read your files, and conversations that follow compatible agent versionsWhat's new

Managed Cloud Get Started
All integrations
Infrastructuretype: ollama

Ollama

Run models locally with Ollama and keep prompts and documents on your own hardware, for on-premises and air-gapped setups.

ConfigurationDocs ↗

What you get

  • An LLM task of type ollama that talks to your Ollama server
  • No provider API key and no data sent to a hosted model
  • A compose overlay that starts Ollama next to EDDI
  • In-process Jlama as an alternative for small quantized models

How it works in EDDI

  1. Start EDDI with the Ollama compose overlay, or run Ollama yourself and note its URL.
  2. Pull a model into Ollama. Use the model tag exactly as ollama list shows it.
  3. Set the task type to ollama with the baseUrl and model.
  4. Inside a container, localhost is the container. Use the compose service name or host.docker.internal.

Configuration

Taken from the EDDI documentation. Placeholders such as your-agent-id are yours to replace.

Start EDDI with Ollama on the same networkBrings Ollama up on the same network as http://ollama:11434.
docker compose -f docker-compose.yml -f docker-compose.ollama.yml up -d
langchain.jsonAn Ollama task from the EDDI docs.
{
  "tasks": [
    {
      "actions": ["send_message"],
      "id": "ollamaChat",
      "type": "ollama",
      "description": "Ollama local model chat",
      "parameters": {
        "baseUrl": "http://ollama:11434",
        "model": "llama3.2:3b",
        "timeout": "120000",
        "think": "false",
        "systemMessage": "You are a helpful assistant",
        "addToOutput": "true"
      }
    }
  ]
}

Governance

  • Prompts, retrieved context and answers stay on your network, which is the point of a local model.
  • Approvals, behavior rules and the audit ledger apply the same as with a hosted model.
  • No credential is involved, so there is nothing to put in the vault for the model itself.

Good to know

  • A fully air-gapped setup needs a local Ollama or a pre-seeded Jlama cache, because EDDI cannot download a model with no network.
  • Local models vary in tool-calling and JSON reliability. Test the behavior you depend on.
Get startedDocs ↗

Need it built and run for you?

EDDI Solutions builds and maintains connectors, runs EDDI as a managed or co-operated instance, and supports you in your own cloud or on-premises.

Talk to EDDI Solutions ↗

Related integrations

Data and RAGChromaChroma as the vector store for an EDDI knowledge base. The lowest-friction option when you have no existing infrastructure to reuse.InfrastructureKubernetes and HelmRun EDDI on any conformant Kubernetes cluster with a single-file quickstart, a Helm chart or Kustomize overlays, with MongoDB or PostgreSQL.ChannelsOpen WebUIEDDI serves an OpenAI-compatible /v1 API that presents deployed agents as models, so Open WebUI, the OpenAI SDKs, LangChain and LiteLLM can talk to them.

Configuration and behavior on this page follow the EDDI documentation, checked on 2026-10-14. Field names and defaults can change between releases, so confirm them in the docs for your version.