PowerPlatformConnectors/certified-connectors/Eden AI

Eden AI

Enhance your apps and workflows with the power of Eden AI! Integrate advanced AI features such as text analysis (sentiment analysis, keyword extraction, text moderation, etc.) , image recognition (object detection, face detection, explicit content detection, etc.), document parsing (invoice, receipt, IDs, resumes, etc.), audio processing, machine translation, generative AI (text generation, image generation, speech generation) and more.

Publisher: Eden AI

Prerequisites

You will need to create your Eden AI account for free.

Obtaining Credentials

Once you have sign up, get your API key from your account:

When you create a connection to Eden AI, do not forget the keyword Bearer before your API key:

Supported Operations

The connector supports the following operations:

Action	Description
Convert Text into Speech	Converts normal language text into speech.
Detect Explicit Content in Images	Detects adult explicit content in images, that is generally inappropriate for people under the age of 18 and includes nudity, sexual activity and pornography...
Text Generation	Generates text based on a given prompt.
Generate Chat Responses	Generates human-like responses to various inputs and queries.
Extract Topic from Text	Extracts general topics in a text.
Extract Keywords from Text	Extracts the most important words and expressions in a text.
Extract Named Entities in Text	Identifies named entities in a text and classifies them into predefined categories.
Anonymize Images	Anonymizes an image by bluring sensitive parts (faces, car plates, etc.)
Detect Faces in Images	Identifies human faces in digital images: position, sex, age, smile, glasses, etc.
Image Generation	Generates compelling images based on a given prompt.
Translate Text into another Language	Translates a text into another language using rules, statics or ML technics.
Moderate Texts	Moderates a text by detecting explicit content.
Summarize a Text	Extracts the most important sentences from a text in order to create a smaller version of the text.
Detect Language of Text	Identifies the specific linguistic features that are unique to each language.
Identify General Sentiment of a Text	Identifies general sentiment of a text and returns positive, negative or neutral.
Extract Information in Invoices	Extracts the data in contains (items, prices, addresses, vendor name, etc.) to automate the invoice processing.
Extract Information in Resumes	Extracts structured data (name, job list, education, skills) to automate the resume processing.
Extract Informations in Identity Documents	Extracts structured information in identity documents (passports, identity cards, driver license, etc.).
Extract Information in Receipts	Extracts structured information like products purchased, quantity, price, date and VAT from receipts.

Convert Text into Speech

Converts normal language text into speech.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• text: Enter the text you want to convert into audio.
• option: Enter MALE or FEMALE to choose the voice gender.
• settings: Specify specific models to use for some providers. It can be in the following format: {'google' : 'google_model', 'microsoft': 'microsoft_model'...}. Check the model available here.
• rate: Increase or decrease the speaking rate by expressing a positif or negatif number ranging between 100 and -100 (a relative value as percentage varying from -100% to 100%).
• pitch: Increase or decrease the speaking pitch by expressing a positif or negatif number ranging between 100 and -100 (a relative value as percentage varying from -100% to 100%).
• volume: Increase or decrease the audio volume by expressing a positif or negatif number ranging between 100 and -100 (a relative value as percentage varying from -100% to 100%).
• audio_format: Optional parameter to specify the audio format in which the audio will be generated. By default, audios are encoded in MP3, except for lovoai which use the wav container.
• sampling_rate: The synthesis sample rate (in hertz) for this audio. When this is specified, the audio will be converted either to the right passed value, or to a the nearest value acceptable by the provider.

Detect Explicit Content in Images

Detects adult explicit content in images, that is generally inappropriate for people under the age of 18 and includes nudity, sexual activity and pornography...

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• file: Choose the file you want to analyze.

Text Generation

Generates text based on a given prompt.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• text: Enter your prompt.
• temperature: A value between 0 and 1. Higher values mean the model will take more risks and value 0 (argmax sampling) works better for scenarios with a well-defined answer.
• max_tokens: A value between 1 and 2048. The maximum number of tokens to generate in the completion. The token count of your prompt plus max_tokens cannot exceed the model's context length.
• settings: Specify specific models to use for some providers. It can be in the following format: {'openai' : 'openai_model', 'cohere': 'cohere_model'...}. Check the model available here.

Generate Chat Responses

Generates human-like responses to various inputs and queries.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• text: Enter your prompt.
• chat_global_action: A system message that helps set the behavior of the assistant. For example, 'You are a helpful assistant'.
• temperature: A value between 0 and 1. Higher values mean the model will take more risks and value 0 (argmax sampling) works better for scenarios with a well-defined answer.
• max_tokens: A value between 1 and 2048. The maximum number of tokens to generate in the completion. The token count of your prompt plus max_tokens cannot exceed the model's context length.
• settings: Specify specific models to use for some providers. It can be in the following format: {'openai' : 'openai_model', 'ibm': 'ibm_model'...}. Check the model available here.

Extract Topic from Text

Extracts general topics in a text.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• text: Tap or paste the text you want to analyze.

Extract Keywords from Text

Extracts the most important words and expressions in a text.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• text: Tap or paste the text you want to analyze.

Extract Named Entities in Text

Generates human-like responses to various inputs and queries.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• text: Tap or paste the text you want to analyze.

Anonymize Images

Anonymizes an image by bluring sensitive parts (faces, car plates, etc.)

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• file: Choose the file you want to analyze.

Detect Faces in Images

Identifies human faces in digital images: position, sex, age, smile, glasses, etc.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• file: Choose the file you want to analyze.

Image Generation

Generates compelling images based on a given prompt.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• text: Tap or paste the text you want to analyze.
• resolution: Enter the resolution, it can only be 256x256, 512x512 or 1024x1024.
• num_images: Enter the number of images you want.

Translate Text into another Language

Translates a text into another language using rules, statics or ML technics.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• text: Tap or paste the text you want to analyze.
• source_language: Enter the language of your text. Check languages supported here.
• target_language: Enter the language of the output text. Check languages supported here.

Moderate Texts

Moderates a text by detecting explicit content.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• text: Tap or paste the text you want to analyze.

Summarize a Text

Extracts the most important sentences from a text in order to create a smaller version of the text.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• text: Tap or paste the text you want to analyze.
• output_sentences: Enter the sentence number of the summary.
• settings: Specify specific models to use for some providers. It can be in the following format: {'google' : 'google_model', 'ibm': 'ibm_model'...}. Check the model available here.

Detect Language of Text

identify the specific linguistic features that are unique to each language.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• text: Tap or paste the text you want to analyze.

Identify General Sentiment of a Text

Identifies general sentiment of a text and returns positive, negative or neutral.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• text: Tap or paste the text you want to analyze.

Extract Information in Invoices

Extracts the data in contains (items, prices, addresses, vendor name, etc.) to automate the invoice processing.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• file: Choose the file you want to analyze.

Extract Information in Resumes

Extracts structured data (name, job list, education, skills) to automate the resume processing.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• file: Choose the file you want to analyze.

Extract Informations in Identity Documents

Extracts structured information in identity documents (passports, identity cards, driver license, etc.).

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• file: Choose the file you want to analyze.

Extract Information in Receipts

Extracts structured information like products purchased, quantity, price, date and VAT from receipts.

• providers: Enter the selected providers seperated by a coma. Check the providers available here.
• language: Check languages supported here.
• file: Choose the file you want to analyze.

Known Issues and Limitations

Here are some issues you could face while using the connector:

• Error 400: Appears when you make a bad request, it means that you wrongly enter the parameters. Check how to configure them correctly with our doc.
• Error 403: Appears when you wrongly enter your API Key. When you create a connection, don't forget the keyword Bearer before your API_Key: Bearer API_KEY

If you have any interrogation about our APIs, check our documentation or feel free to contact us.

Deployment Instructions

Please use the instructions on Microsoft Power Platform Connectors CLI to deploy this connector as a custom connector in Microsoft Power Automate or Power Apps.