ollama_chat
Chat with a model using conversation messages. Supports system messages, multi-turn conversations, tool calling, and generation options.
LABSCO SUMMARY
Model management stops being a terminal errand the agent has to ask you to run.
Inference and library administration sit in the same tool list, so an agent that needs a model it does not have can pull it, check what is loaded, run the task and delete it afterwards inside one session. The two web tools are the only ones that reach off the machine, and they are the only ones that need a key, which keeps the boundary easy to reason about. AGPL-3.0 is the one detail to settle before this goes inside something you ship.
What it is
A server over a local Ollama instance covering inference, embeddings and the model library, plus Ollama's hosted web search and fetch.
What you get
- Multi-turn chat with system messages, tool calling and generation options, or a single-turn completion when the conversation shape is not needed
- Embeddings returned as numerical vectors for text you pass in
- The local library administered end to end: list what is installed, show a model's modelfile, parameters and architecture, pull, push, copy and delete
- Which models are loaded in memory right now, separately from which are on disk
- New models created from an existing one with their own system prompt, template and license
- Web search and page fetch through Ollama's hosted APIs, returning a page's title, content and links to ground an answer in current material
Requirements
An Ollama instance the server can reach, with the models you intend to call already pulled into it. OLLAMA_API_KEY is needed by ollama_web_search and ollama_web_fetch, which call Ollama's hosted APIs; nothing else in the list asks for it. Licensed AGPL-3.0.
Setup effort
One command — npm install -g ollama-mcp
USE CASES
01
Getting a model onto the machine before it is needed
Pull the weights, confirm they landed, and read what the model actually is — its modelfile, parameters and architecture — before building anything on it.
02
Keeping a sensitive task on local hardware
Run the work against an installed model instead of a hosted one — multi-turn where the conversation matters, single-turn completion where it does not.
03
Making a house-style variant of a base model
Create the variant with its own system prompt and template, keep a copy under a second name before you change it further, and push it when it is worth sharing.
04
Freeing memory and disk after an experiment
Check what is still resident in memory as opposed to merely installed, then remove the models the experiment left behind.
05
Answering from a current page rather than from training data
Search the web through Ollama's hosted API and fetch the page that looks right, taking back its title, content and links. Both calls need OLLAMA_API_KEY.
13 tools
Counted once with only startup setup. Other versions, modes or optional credentials may expose a different list.
ollama_copy
Copy a model. Creates a duplicate of an existing model with a new name.
ollama_create
Create a new model with structured parameters. Allows customization of model behavior, system prompts, and templates.
ollama_delete
Delete a model from local storage. Removes the model and frees up disk space.
ollama_embed
Generate embeddings for text input. Returns numerical vector representations.
ollama_generate
Generate completion from a prompt. Simpler than chat, useful for single-turn completions.
ollama_list
List all available Ollama models installed locally. Returns model names, sizes, and modification dates.
ollama_ps
List running models. Shows which models are currently loaded in memory.
ollama_pull
Pull a model from the Ollama registry. Downloads the model to make it available locally.
ollama_push
Push a model to the Ollama registry. Uploads a local model to make it available remotely.
ollama_show
Show detailed information about a specific model including modelfile, parameters, and architecture details.
ollama_web_fetch
Fetch a web page by URL using Ollama's web fetch API. Returns the page title, content, and links. Requires OLLAMA_API_KEY environment variable.
ollama_web_search
Perform a web search using Ollama's web search API. Augments models with latest information to reduce hallucinations. Requires OLLAMA_API_KEY environment variable.
BEFORE YOU INSTALL
What to know first
Tools you get
13 tools when we connected. ollama_web_search and ollama_web_fetch require an Ollama Cloud OLLAMA_API_KEY, so a purely local install has eleven working tools. The project's 14-tool figure is stale, and ollama_your_tool is example scaffolding, not a tool.
Requires
- npx on PATH
- Node.js v16 or later
Where it runs
All 3 documented clients run the server on your machine.
CONTEXT COST
Light Standing cost
Rated by the size of the 13 tool definitions that ride along on every turn, not by how many there are
INSTALLATION
Connect your client
Server setup
TERMINAL
npm install -g ollama-mcpMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"ollama": {
"command": "npx",
"args": [
"-y",
"ollama-mcp"
]
}
}
}~/Library/Application Support/Claude/claude_desktop_config.json on macOS
Cline
CLINE · JSON
{
"mcpServers": {
"ollama": {
"command": "npx",
"args": [
"-y",
"ollama-mcp"
]
}
}
}cline_mcp_settings.json
Any MCP client — hybrid cloud
ANY MCP CLIENT — HYBRID CLOUD · JSON
{
"mcpServers": {
"ollama": {
"command": "npx",
"args": [
"-y",
"ollama-mcp"
],
"env": {
"OLLAMA_HOST": "http://127.0.0.1:11434",
"OLLAMA_API_KEY": "<YOUR_OLLAMA_API_KEY>"
}
}
}
}
