Usage
Learn how to use LLM Vision
Last updated
Learn how to use LLM Vision
Last updated
LLM Vision exposes four actions (previously called services):
image_analyzer: Analyzes image files, image and camera entities
video_analyzer: Analyzes video files and Frigate events
stream_analyzer: Records and analyzes camera entities
data_analyzer: Updates a sensor value based on charts, visualizations etc.
Below are all parameters for all actions. Not all actions accept all parameters. See the Actions' page for more information.
| Parameter | Required | Description | Default | Valid Values | Actions |
|---|---|---|---|---|---|
provider
Yes
The AI provider configuration
Any configuration
All
model
No
Model used for processing the image(s).
gpt-4o-mini
See 'Choosing the right model'
All
message
Yes
The prompt to send along with the image(s).
String
All
remember
No
Remember the analyzed event
false
true, false
Image, Video, Stream
sensor_entity
Yes
Sensor to update
sensor, input_datetime, input_number, input_text
Data
image_file
No*
The path to the image file(s). Each path must be on a new line.
Valid path to an image file
Image
image_entity
No*
An alternative to image_file for providing image input.
any image or camera entity
Image, Stream
video_file
No*
The path to the video file(s). Each path must be on a new line.
Valid path to an video file
Video
event_id
No*
Event ID from Frigate. Each id must be on a new line.
a Frigate event id
Video
max_frames
No
How many frames to analyze. Will pick the most relevant frames (most motion)
3
Integer 1-10
Video, Stream
duration
Yes
For how many seconds to capture and analyze stream.
10
Integer between 1 and 300
Video, Stream
include_filename
Yes
Whether to include the filename in the request.
false
true, false
All
target_width
No
Width to downscale the image to before encoding.
1280
Integer between 512 and 3840
All
detail
No
Level of detail to use for image understanding.
auto
auto, low, high
All
max_tokens
Yes
The maximum number of response tokens to generate.
100
Integer between 10 and 1000
All
temperature
Yes
Randomness of the output.
0.5
Float between 0.0 and 1.0
All
expose_images
No
Save all analyzed frames to /www/llmvision
false
true, false
Image, Video, Stream