Ollama Vision for Home Assistant (Marketplace integration)

Installs: 188 sharing analytics
GitHub stars: 22
Connection: Works locally
Upkeep: Updated in the last 6 months
Version: 1.0.10, October 4, 2026
Category: AI and voice, AI models
Kind: Marketplace integration
Brand: Ollama
Needs: Needs the Marketplace (HACS on 2026.10 and older)
Domain: ollama_vision
Source: github.com/remimikalsen/ollama_vision
Issues: Report an issue

Independent directory, made by the developer of Vulpo. Not affiliated with Home Assistant or the Open Home Foundation.

About

Ollama Vision is an integration that describes images using a vision-capable model on a local Ollama server. It adds an action that sends an image for analysis and writes the result to a sensor, which is created or updated with the latest description. An optional second text model can elaborate on the description, and each model has its own host, model name and keep-alive setting. Images can come from web addresses, Frigate, Google Nest cameras or local files, and prompts can be customized. An event fires after each image is analyzed, so automations can pick up the result, for example to send it to a phone. Several configurations can be added, each appearing as a device with its own sensors, and setup is done in the UI.

Automatic summary of the README and code of remimikalsen/ollama_vision. Something wrong, or are you the maintainer and want your own text here? Get in touch

Supports

Works locally Talks to your devices on your own network. No vendor cloud in between.
From the Marketplace Installed through the Marketplace, formerly HACS.
Actively maintained Its author released or pushed a change in the last six months.

Provides

Sensors

Setting it up

Install it from the Marketplace and restart Home Assistant, then add it under Settings, Devices and services.

Connect to Ollama Vision

Name
Vision Host
Vision Port
Vision Model
Vision Model Keep-Alive (-1 for indefinite)
Enable Text Model

Text Model Configuration

Text Model Host
Text Model Port
Text Model
Text Model Keep-Alive (-1 for indefinite)

After setup

Vision Host
Vision Port
Vision Model
Vision Model Keep-Alive (-1 for indefinite)
Enable Text Model
Text Model Host
Text Model Port
Text Model
Text Model Keep-Alive (-1 for indefinite)

Texts from the code of remimikalsen/ollama_vision.

Actions

Analyze ImageSend an image to Ollama for analysis and create a sensor with the result. Optionally elaborate the description with a text model.Fields: Image URL, Vision Prompt, Image Name, Configuration, Use Text Model, Text Prompt

From the README

This integration enables you to analyze images using a local Ollama server running a vision-enabled LLM. Optionally, it can also use a specialized text model to generate more elaborate text. For example, you could add a cheeky or “roast-style” introduction for people captured on a security camera.
You can for example use external images reachable over http/https, Frigate, Google Nest cameras or even local files as the image source for analysis. Have your local LLM describe images and send those descriptions to your phone, or use them in automations however you like.
Connect to one or more Ollama servers with vision-capable models
Analyze images with customizable prompts
Optionally enhance descriptions using a specialized text model
Dynamically create/update sensors with the latest image descriptions
Fire an event when each image is analyzed, for downstream automations

From the README of remimikalsen/ollama_vision, MIT License. Shortened.