Enables or disables the AI Translation module for the selected store view. Set to Yes to

activate all AI translation features.

Select the AI provider used for translations. Currently supports OpenAI (ChatGPT). The AI

model list updates automatically based on the selected provider and API key.

Enter your OpenAI API key generated from the OpenAI platform. This key is required to

authenticate and use AI translation services.

Select the AI model to be used for translations. Available options include:

• GPT-4.5 (Strongest) – Highest translation accuracy, best for complex, SEO-critical, and

brand-sensitive content.

• GPT-4.1 (Best for Coding) – Optimized for structured and technical content such as product

specifications and attributes.

• GPT-4.1 Mini (Fast) – Faster responses with balanced quality and cost, ideal for bulk

translations.

• GPT-4.1 Nano (Cheapest) – Most cost-effective option, suitable for large catalogs with

basic translation needs.

• OpenAI o3 (Reasoning) – Enhanced reasoning capabilities, useful for context-heavy or

nuanced translations.

• OpenAI o3 Pro – Advanced reasoning model offering higher consistency and contextual

understanding.

• OpenAI o4 Mini – Lightweight model designed for efficient, high-volume translations.

• GPT-4o (Multimodal) – Supports advanced contextual understanding and multimodal

capabilities, suitable for rich content translations.

• GPT-4o Mini (Efficient) – Optimized for speed and cost efficiency while maintaining strong

translation quality

.• GPT-3.5 Turbo (Legacy) – Legacy model for basic translation use cases and backward

compatibility.

The available model list dynamically updates based on the selected AI service and API key.

Controls the randomness and creativity of translations. Lower values (0) produce precise and literal

translations, while higher values (up to 2) allow more creative wording. Default recommended value

is 0.7.

Test the connection between Magento and the AI service using the provided API key and selected

model to ensure the configuration is valid.

Enables or disables the AI Translation module for the selected store view. Set to Yes to

activate AI-based translations.

Select Google Gemini as the AI service provider for translations. The available AI model

list updates automatically based on the selected provider and API key.

Enter your Google Gemini API key generated from Google AI Studio. This key is required

to authenticate and use Gemini models for translation.

Select the Google Gemini model to be used for translations. Available options include:

• Gemini 2.5 Flash – Fast and efficient model with strong quality, suitable for bulk

translations.

• Gemini 2.5 Pro – High-quality model with advanced reasoning, ideal for SEO-critical

and brand-sensitive content.

• Gemini 2.0 Flash – Balanced performance model for general translation use cases.

• Gemini 2.0 Flash-Lite – Lightweight and cost-efficient option for large catalogs.

• Gemini Flash Latest – Always points to the latest stable Flash model released by

Google.

• Gemini Flash-Lite Latest – Latest lightweight Flash-Lite model optimized for cost

and speed.

• Gemini Pro Latest – Latest stable Pro model with the best overall translation quality

and context handling.

• Gemini 2.5 Flash-Lite – Cost-optimized variant of Gemini 2.5 Flash for high-volume

translation workloads.

• Gemini 2.0 Flash Experimental – Experimental version with early access to new

capabilities (not recommended for production)

.• Gemini 2.0 Flash 001 – Fixed version of Gemini 2.0 Flash for consistent and

repeatable outputs.

• Gemini 2.0 Flash-Lite 001 – Fixed lightweight Flash-Lite version for predictable

performance.

• Gemini 2.0 Flash-Lite Preview 02-05 – Preview release for testing upcoming Flash

-Lite improvements.

• Gemini 2.0 Flash-Lite Preview – Early preview model intended for evaluation purposes

only.

• Gemini Experimental 1206 – Experimental Gemini model with cutting-edge features,

recommended only for testing.

• Gemini 2.5 Flash Preview Sep 2025 – Preview of upcoming Gemini 2.5 Flash release.

• Gemini 2.5 Flash-Lite Preview Sep 2025 – Preview version of the Flash-Lite model for

early access testing.The available model list dynamically updates based on the selected

AI service and API key.

Controls the randomness and creativity of translations. Lower values (0) produce more

literal translations, while higher values (up to 2) allow more creative phrasing. Default

recommended value is 0.7.

Tests the connection between Magento and Google Gemini using the provided API key

and selected model to ensure the configuration is valid.

Enables or disables the AI Translation module for the selected store view. Set to

Yes to activate AI-powered translations.

Select Anthropic Claude as the AI service provider for translations. The AI model

list updates automatically based on the selected provider and API key.

Enter your Anthropic Claude API key generated from the Anthropic Console. This

key is required to authenticate and use Claude models.

Select the Claude model to be used for translations. Available options include:

• Claude Sonnet 4 (Latest – Smart & Efficient) – Balanced, intelligent, and optimized

for most translation use cases.

• Claude Opus 4 (Most Capable) – Most advanced Claude model, recommended for

complex, nuanced, and brand-sensitive content.

• Claude 3.5 Sonnet – Strong balance between speed and quality, ideal for product

descriptions, CMS content, and SEO translations.

• Claude 3.5 Haiku (Fast & Efficient) – Lightweight and fastest option, suitable for bulk

or short-form translations.

• Claude 3 Opus – High reasoning capability within the Claude 3 family, suitable for

detailed and context-heavy translations.

• Claude 3 Sonnet – General-purpose model offering a balance of performance and

accuracy for everyday translation needs.

• Claude 3 Haiku – Cost-efficient model for simpler content and large-scale translations.

• Claude 2.1 (Legacy) – Older stable release maintained for backward compatibility.

• Claude 2.0 (Legacy) – Legacy model with basic capabilities, retained for historical and

compatibility purposes.

The available model list dynamically updates based on the selected AI service and API key.

Controls translation creativity and variation. Lower values (0) produce precise and literal translations,

while higher values (up to 2) allow more expressive phrasing. Default recommended value is 0.7.

Tests the connection between Magento and Anthropic Claude using the provided API key and

selected model to ensure the configuration is valid.

Specifies the API endpoint used for sending translation and content generation requests to OpenAI. By default, this is set to

https://api.openai.com/v1/chat/completions. You can override this value if you are using a proxy server,

API gateway, or custom OpenAI-compatible deployment.

Defines the API endpoint used to fetch the list of available OpenAI models. The default endpoint is

https://api.openai.com/v1/models. Customizing this is useful when routing requests through a proxy or

using a self-hosted or enterprise-managed OpenAI-compatible service.

Defines the base API endpoint used to send translation and content generation requests to

Google Gemini. By default, this is set to

https://generativelanguage.googleapis.com/v1beta/models/. The selected

model name is automatically appended to this URL when making requests. This field can be

customized when using proxy servers or custom Gemini deployments.

Specifies the endpoint used to retrieve the list of available Google Gemini models. The default

value is https://generativelanguage.googleapis.com/v1beta/models. Customizing

this endpoint is useful for advanced configurations such as routing requests through a proxy or

enterprise-managed API gateway.

Specifies the API endpoint used to send translation and content generation requests to Anthropic

Claude. By default, this is set to https://api.anthropic.com/v1/messages. You can

override this value if you are using a proxy server, API gateway, or custom Anthropic-compatible

deployment.

Defines the endpoint used to retrieve the list of available Claude models. The default endpoint

is https://api.anthropic.com/v1/models. Customizing this endpoint is useful for

advanced configurations such as routing requests through a proxy or enterprise-managed

API infrastructure.

Defines the maximum number of tokens allowed in a single AI response. This setting controls

how long and detailed the translated content can be. The default value is 3900, which is suitable

for most product descriptions, CMS pages, and SEO content. Increasing this value allows longer

translations but may increase API usage and cost.

System-friendly unique code used internally to identify the prompt template (e.g.,

english_arabic_tone).

Human-readable name of the prompt template that describes its purpose (e.g.,

English to Arabic Tone).

Specifies the attribute or content type this prompt applies to, such as short_description,

description, meta_title, etc.

Defines the Magento entity where this prompt is applied, such as product, category,

or CMS page/block.

Indicates whether the prompt template is active. Only enabled templates are used

during translation.

Determines whether this prompt template is the default one for the selected entity and

content type. Only one default template can exist per combination.

Controls the priority of the prompt template. Lower values are processed first when

multiple templates are available.

A unique system identifier for the prompt template. Use only lowercase letters, numbers,

and underscores (e.g., english_arabic_tone). This code is used internally and must

be unique.

A human-readable name describing the purpose of the prompt template, such as English

to Arabic Tone. This helps admins easily identify templates.

Specifies which content attribute this prompt applies to, such as Product Short Description,

Product Description, Meta Title, Meta Description, etc.

Defines the Magento entity for which this prompt is used. Common values include product,

category, cms_page, or pagebuilder.

If enabled, this template becomes the default for the selected entity type + content type

combination. Only one default template is allowed per combination.

Sets the maximum number of tokens the AI can generate for this specific content type. This

helps control output length and API usage.

Determines the priority of the template. Lower values are evaluated first when multiple

templates are available.

Optional helper text or sample content used to quickly test the prompt behavior before saving. This

does not affect live translations.

Runs a test translation using the sample content and current prompt configuration to preview the

output.

The core instruction sent to the AI model. This defines translation rules, tone, language style, SEO

behavior, and how variables are used. The prompt supports dynamic placeholders such as

{{source_text}}, {{target_language}}, and entity variables. This field is mandatory.

Select the AI provider used for translation (e.g., Google Gemini, OpenAI, Anthropic Claude). This field

is mandatory.

Choose the AI model for translation (e.g., Gemini 2.5 Flash). Click Refresh to reload available models

based on the selected service and API key.

Displays the source language and locale of the content being translated

(e.g., English – United States en_US).

Select the target store view where the translated content will be saved. This determines the target

language and locale. This field is mandatory.

Select the prompt template that defines translation rules, tone, and formatting. The selected

template controls how the AI processes the content.

Shows the final prompt sent to the AI, including dynamically injected variables such as

{{target_language}} and the source content. This field is read-only for preview purposes.

Displays the supported variables used in the prompt, such as {{source_language}},

{{target_language}}, and {{source_text}}.

Current state of the translation. Possible values: Pending (awaiting review), Approved &

Published (saved to target store), or Rejected (not published).

Select the prompt template that defines translation tone, language rules, SEO

behavior, and brand terminology (e.g., English to Arabic Tone).

Choose the product attribute to be translated (e.g., Product Short Description,

Description, Meta Title).

Store view from which the original content will be taken

(typically Default Store View).

When enabled, products with empty source attributes are skipped during

translation.

When enabled, existing translations in the target store view will be replaced

with new AI-generated content.

Number of products translated per batch (1–50). Smaller batch sizes provide

better stability and error handling.

Select the prompt template that defines tone, language

rules, SEO behavior, and brand terminology (e.g., English to Arabic Tone).