Ollama MCP Server

Enables seamless integration between local Ollama LLM instances and MCP-compatible applications for advanced task management.

Local serverstdioPython 6

What is the Ollama MCP server?

If you already use Ollama, the ollama mcp server is the piece that lets your assistant work with it directly. Enables seamless integration between local Ollama LLM instances and MCP-compatible applications for advanced task management.

What the server does

このMCPサーバーは、ローカルのOllama LLMインスタンスとMCP互換アプリケーションの間でシームレスな統合を可能にし、高度なタスク分解、評価、ワークフロー管理を提供します。

Available tools

The toolset is worth reading before you wire it up, because it tells you what the integration is really for:

  • task — //** - 個別のタスクにアクセスするためのURIスキーム
  • result — //** - 評価結果にアクセスするためのURIスキーム
  • model — //** - 利用可能なOllamaモデルにアクセスするためのURIスキーム
  • decompose-task — 複雑なタスクを管理しやすいサブタスクに分解
  • evaluate-result — 指定された基準に対してタスク結果を分析
  • add-task — 必須パラメータ: name (文字列), description (文字列)
  • run-model — 必須パラメータ: model (文字列), prompt (文字列)
  • MacOS — The MacOS tool exposed by this server
  • Windows — The Windows tool exposed by this server

Installation

@modelcontextprotocol/inspector on npm is all you need. Most clients run it directly, so configuration is a few lines and a restart.

Credentials and setup notes

Configuration is passed through the environment: OLLAMA_HOST, UV_PUBLISH_TOKEN. Treat anything key-shaped as a real credential — scope it to the minimum the server needs, and rotate it if it ever lands in a shared config.

Where it fits

Plenty of planning and project tracking servers cover similar ground. The differences that matter in practice are scope of access and how much setup stands between you and a working tool call. Ollama's toolset — task, result, model and 6 more — is a fair guide to whether it matches your workflow. It is maintained by NewAITees; worth a glance at recent repository activity before you build anything load-bearing on it.

We check each listing at SyncDev against the project's documentation before it goes live — if something here drifts out of date, it is a bug worth reporting.

Worth knowing first

  • It runs with your machine's permissions. That is convenient and also the reason to think about what you point it at before you approve a tool call.
  • With 9 tools registered it takes up a noticeable share of the context window; turn it off in projects that never touch Ollama.
  • Missing credentials fail quietly in some clients — if no tools show up, check the environment block first.
  • Keep per-call confirmation enabled while you learn its behaviour; it is the cheapest safeguard you have.

Available tools

ToolWhat it does
task//** - 個別のタスクにアクセスするためのURIスキーム
result//** - 評価結果にアクセスするためのURIスキーム
model//** - 利用可能なOllamaモデルにアクセスするためのURIスキーム
decompose-task複雑なタスクを管理しやすいサブタスクに分解
evaluate-result指定された基準に対してタスク結果を分析
add-task必須パラメータ: name (文字列), description (文字列)
run-model必須パラメータ: model (文字列), prompt (文字列)
MacOSThe MacOS tool exposed by this server.
WindowsThe Windows tool exposed by this server.

How to install the Ollama MCP server

"mcpServers": {
    "ollama-MCP-server": {
      "command": "uv",
      "args": [
        "--directory",
        "/path/to/ollama-MCP-server",
        "run",
        "ollama-MCP-server"
      ],
      "ENV":["model":"deepseek:r14B"]
    }
  }

Configuration as documented by the project. Restart the client after saving.

Configuration

VariableDescriptionRequired
OLLAMA_HOSTEndpoint or connection string the server talks to.Optional
UV_PUBLISH_TOKENCredential the server authenticates with.Yes

Example prompts to try

  • Use Ollama to task.
  • Use Ollama to result.
  • Use Ollama to model.

Frequently asked questions

Key features include complex task decomposition, result validation, Ollama model management, standardized communication via MCP, and detailed error handling.