Ask Ollama

Name: Ask Ollama
Author: Lykhoyda

by Lykhoydacommunityv0.4.0

MCP servers that let an AI client send prompts to multiple LLM providers including Gemini, Codex, and local Ollama through standard MCP. It also includes a Claude Code plugin for multi-provider code review, brainstorming, comparison, and a pre-commit review hook.

From the registry: Bridge Claude with local Ollama LLMs for private AI-to-AI collaboration — no API keys, fully local

GitHub Repository

Install & configure

Prerequisites

required
Install Node.js
Node.js v20.0.0 or higher is required to run the MCP server via npx.
https://nodejs.org/
required
Install and run Ollama
Ollama must be installed and running locally for ask-ollama-mcp to work.
https://ollama.com/
required
Pull an Ollama model
Pull at least one local model. The README example uses qwen2.5-coder:7b, which is also the documented default model.
```
$ ollama pull qwen2.5-coder:7b
```

Install in your client

Paste into the AI you're chatting with — it will figure out the rest

Please install the `ask-ollama` MCP server into my current AI client (that's you).

Required prerequisites (do these first if not already done):
- **Install Node.js** — Node.js v20.0.0 or higher is required to run the MCP server via npx. (https://nodejs.org/)
- **Install and run Ollama** — Ollama must be installed and running locally for ask-ollama-mcp to work. (https://ollama.com/)
- **Pull an Ollama model** — Pull at least one local model. The README example uses qwen2.5-coder:7b, which is also the documented default model. Run: `ollama pull qwen2.5-coder:7b`

Canonical MCP server config (stdio transport):
- command: `npx`
- args: ["-y","ask-ollama-mcp"]

Note: The README provides a dedicated Ollama MCP server package. Preferred portable install path is via npx using the stdio transport. No environment variables are documented for ask-ollama-mcp. The server relies on a local Ollama instance and defaults to model qwen2.5-coder:7b, with qwen2.5-coder:1.5b as fallback if not found.

Add this MCP server to my current client's config in the correct format for you. If you need secrets or credentials I haven't provided, ASK me — do not invent values or leave raw placeholders. After adding it, tell me how to verify the server is connected.

Note: The README provides a dedicated Ollama MCP server package. Preferred portable install path is via npx using the stdio transport. No environment variables are documented for ask-ollama-mcp. The server relies on a local Ollama instance and defaults to model qwen2.5-coder:7b, with qwen2.5-coder:1.5b as fallback if not found.

Tools (10)

ask-gemini

Send prompts to Gemini CLI with @ file syntax and progressive streaming output.

ask-gemini-edit

Return structured OLD/NEW code edit blocks from Gemini.

fetch-chunk

Retrieve chunks from cached large Gemini responses.

ask-codex

Send prompts to Codex CLI with session resume support.

ask-ollama

Send prompts to local Ollama for fully private responses.

ask-llm

Unified orchestrator that routes a prompt to a chosen provider or fans out to all installed providers.

multi-llm

Send the same prompt to multiple providers in parallel and return per-provider responses and usage.

get-usage-stats

Return per-session token totals, fallback counts, and provider/model breakdowns.

diagnose

Run read-only self-diagnosis for Node version, PATH resolution, and provider CLI presence and versions.

ping

Test the MCP connection and verify setup.

Packages

ask-ollama-mcp

npmv0.4.0transport: stdio

Details

Language: TypeScript
License: MIT
Released: Jun 15, 2026
Last commit: Jul 4, 2026
Registry name: io.github.Lykhoyda/ask-ollama

Related Servers

Claude Flow

AI orchestration with hive-mind swarms, neural networks, and 87 MCP tools for enterprise dev.

65.5k

Codebase Memory Mcp

Persistent codebase knowledge graph. Survives session restarts and context compaction.

33.8k

Serena

A powerful toolkit for coding, providing semantic retrieval and editing capabilities.

26.7k