Skip to content

Multimodal

Image, video, speech, and diagram generation tools.

← Prebuilt Tools overview

ToolExtraDescription
generate_image_stabilitybaseGenerate images with Stability AI.
image_readerbaseAnalyze local image files.
generate_imagebaseGenerate images with Amazon Bedrock.
nova_reelsbaseCreate short videos with Nova Reels on Bedrock.
speakbaseText-to-speech via macOS say or Amazon Polly.
diagramdiagramRender cloud architecture and UML diagrams.
chat_videotwelvelabsQ&A about video content via TwelveLabs.
search_videotwelvelabsSemantic search over video libraries.

generate_image_stability

Generate images with Stability AI.

python
from elsai_tools.generate_image_stability import generate_image_stability
ParameterTypeRequiredDescription
promptstrYesImage prompt
widthintNoImage width
heightintNoImage height
stylestrNoStyle preset
seedintNoRandom seed

image_reader

Analyze local image files.

python
from elsai_tools.image_reader import image_reader
ParameterTypeRequiredDescription
image_pathstrYesPath to image file
promptstrNoAnalysis prompt

generate_image

Generate images with Amazon Bedrock.

python
from elsai_tools.generate_image import generate_image
ParameterTypeRequiredDescription
promptstrYesImage prompt
negative_promptstrNoNegative prompt
widthintNoImage width
heightintNoImage height
seedintNoRandom seed

nova_reels

Create short videos with Nova Reels on Bedrock.

python
from elsai_tools.nova_reels import nova_reels
ParameterTypeRequiredDescription
actionstrYescreate, status, or list
textstrNoVideo prompt
image_pathstrNoSource image
s3_bucketstrNoOutput S3 bucket
fpsintNoFrames per second
dimensionstrNoVideo dimensions

speak

Text-to-speech via macOS say or Amazon Polly.

python
from elsai_tools.speak import speak
ParameterTypeRequiredDescription
textstrYesText to speak
voicestrNoVoice ID
enginestrNoTTS engine

diagram

Render cloud architecture and UML diagrams.

python
from elsai_tools.diagram import diagram

Extra: diagram

ParameterTypeRequiredDescription
diagram_typestrYesDiagram type
nodesList[Dict[str, str]]NoNode definitions
edgesList[Dict[str, Union[str, int]]]NoEdge definitions
output_formatstrNoOutput format (default png)
titlestrNoDiagram title
elementsList[Dict[str, str]]NoUML elements

chat_video

Q&A about video content via TwelveLabs.

python
from elsai_tools.chat_video import chat_video

Extra: twelvelabs

ParameterTypeRequiredDescription
video_idstrYesTwelveLabs video ID
promptstrYesQuestion about the video
conversation_idstrNoConversation ID for follow-ups

search_video

Semantic search over video libraries.

python
from elsai_tools.search_video import search_video

Extra: twelvelabs

ParameterTypeRequiredDescription
querystrYesSearch query
index_idstrNoVideo index ID
max_resultsintNoMax results

Copyright © 2026 Elsai Foundry.