Chat GPT

Ask questions about your analytics and get AI insights on any Matomo report with ChatGPT or any OpenAI-compatible API, plus an optional agent that queries your live reports through the Matomo MCP tools.

Features

Insights on any report

• An Insights button in the report header opens an insight panel that analyses the report you are looking at, then lets you ask follow-up questions.
• The whole report is analysed, not only the rows visible on screen: the plugin sends a compact payload with the rows, the report totals and the names and units of the metrics.
• The analysis follows what you see: the active segment, the period and date, and the compared periods or segments.
• The analysis also follows the table as you left it: the search, the sort, the flattening and up to 5 expanded rows, whose subtables are added to the analysis.
• Works with data tables, evolution graphs, goals, ecommerce, custom dimensions, custom reports, row evolution and the other reports Matomo declares as widgets.
• When a widget has no report data, or the report cannot be loaded (for example a missing access), a clear message is displayed instead of sending the error to the model.
• The panel is an accessible dialog: keyboard navigation, focus kept inside the panel, Escape to close.

Chat

• A ChatGPT page in the main menu, with suggested questions to start a conversation and a "New conversation" button.
• Streaming answers, with an automatic fallback when the endpoint does not stream.
• Auto-scroll that follows the answer while it is written, stops when you scroll up, and a button to jump back to the latest message.
• Markdown answers with scrollable tables and code blocks, and buttons to copy an answer or a code block.
• Enter sends the message, Shift+Enter adds a new line.

Agent mode, with AI Providers and MCP Server

When a provider is connected in AI Providers and the McpServer plugin is enabled, the chat and the insight panel work as an agent:

• The agent queries your live reports, websites, goals, dimensions and segments by itself, for any period, to answer with real figures.
• A timeline shows each Matomo tool used for an answer and its status.
• Tools run inside Matomo with the permissions of the current user: no public URL, OAuth client or extra token is needed.
• Write actions (for example creating a goal, an annotation or a segment) are only possible when McpServer allows write methods, and only after the agent has described the exact change and you have confirmed it explicitly in the conversation. This rule is added outside the editable prompts, so it cannot be removed by a custom prompt.
• When something is missing, the chat recommends the next step, one at a time: activate AI Providers, connect a provider, install, activate or enable MCP Server, allow write methods. Super users get a direct link for each step, other users are asked to contact their Matomo administrator.

The agent mode is optional. Without AI Providers or McpServer, the plugin keeps working with its own settings. With AI Providers but without McpServer, the connected provider answers without tools.

Connection and models

• One key is enough, used in this order: the API key set for the website, then the provider connected in AI Providers, then the general API key of the plugin. Keys are never copied from one place to another.
• When a provider is connected in AI Providers, it takes over: the host, API key and model of the general settings are displayed read-only, and are used again if AI Providers is deactivated. A key set for a website still overrides AI Providers for that website (the agent mode is then not used for that website).
• Works with OpenAI (default) and any OpenAI-compatible endpoint served over HTTPS, such as Azure OpenAI or a self-hosted server. The API key is optional on a custom host.
• The general API key is only sent to the general host: a website using its own host never receives it.
• Preset models, including Latest recommended (the default, which follows the model recommended by each plugin release, currently GPT 6 Astra), or any custom model name.
• When a model is deprecated, retired or not available to your OpenAI account, the chat explains it and links to the settings instead of showing the raw API error.

Prompts

• Default prompts written for analytics: the chat answers as a senior analytics consultant (direct answer, key figures, prioritised recommendations, no invented figures), and insights follow a fixed structure (summary, key figures, notable patterns, recommendations).
• Default prompts of previous versions are upgraded automatically to the new defaults, in the language of each user. Custom prompts are kept.
• A prompt equal to the default is not stored, so future improvements of the defaults reach you. A Reset to default button restores the default prompts.

More

• Interface translated into 13 languages: English, Arabic, Chinese (Simplified and Traditional), Dutch, French, German, Italian, Japanese, Polish, Portuguese, Spanish and Swedish.
• Follows the Matomo light and dark themes.
• Rate limit of 30 AI requests per hour, per user and per website, configurable in config/config.ini.php.
• HTTP API: ChatGPT.getResponse, ChatGPT.getStreamingResponse, ChatGPT.getInsights, ChatGPT.getSiteSettings, ChatGPT.setSiteSettings and ChatGPT.setSystemSettings.

Requirements

• Matomo 6.x (>=6.0.0-b1,<7.0.0-b1)
• PHP 8.1 or higher
• One of: an OpenAI API key, an OpenAI-compatible HTTPS endpoint, or a provider connected in AI Providers
• Optional, for the agent mode: the AI Providers plugin (bundled with Matomo) with a connected provider, and the McpServer plugin from the Marketplace. Write actions also require write methods to be allowed in the McpServer settings (Raw Matomo API tool access).

Installation / Configuration

1. Install and activate ChatGPT from Administration > Platform > Marketplace.
2. As a super user, open Administration > System > ChatGPT:
   • Connection card: host (an HTTPS URL), API key and model. Each card is saved on its own. The saved key is never displayed, and a Delete key button removes it.
   • Privacy card: check Allow sending Matomo data to the AI provider, off by default. Until a super user checks it, nothing is sent to the provider. The card shows where the data goes, and masks e-mail and IP addresses, removes URL query strings and excludes visitor-level data, all three on by default.
   • Prompts card: chat and insight base prompts, with a Reset to default button.
   • If a provider is connected in Administration > System > AI Providers, the connection fields are read-only and AI Providers is used.
3. Optionally, override the settings for a website in Administration > Websites > ChatGPT (website admin access): host, API key, model and prompts, with a Delete key button and a Use the general prompts button. Empty fields use the general settings.
4. For the agent mode, connect a provider in AI Providers, then install, activate and enable McpServer (Administration > System > General settings > McpServer). The chat guides you through each missing step.

5. Optionally, as a server administrator, tune the plugin in a [ChatGPT] section of config/config.ini.php. Every key is optional:

   [ChatGPT]
   ; name shown in the top menu, on the chat page and in the insights panel, 40 characters max
   display_name = "Company AI"
   ; AI requests per user and per website within the window, 0 disables the limit
   rate_limit_requests = 30
   ; length of the window, in seconds
   rate_limit_window = 3600
   

Then open the ChatGPT page in the main menu, or click the Insights button in the header of a report.

Privacy and data

• Insights send the data of the report you are looking at (labels, metrics and totals), its period, segment and comparisons, and the conversation, to the configured endpoint: OpenAI by default, your custom host, or the provider connected in AI Providers.
• The chat sends your messages and the prompt. In agent mode, the results of the Matomo tools the agent calls are also sent to the provider.
• Nothing is sent to the AI provider until a super user checks Allow sending Matomo data to the AI provider in Administration > System > ChatGPT, Privacy card. It is off by default: until then, the chat and the insights tell users to ask a super user to allow it.
• Before sending, e-mail and IP addresses are replaced with [email] and [ip], and URL query strings are removed, in the report data and in the results of the Matomo tools. Both are on by default and can be turned off in the same Privacy card.
• Visitor-level data (Visits Log, visitor profiles, real-time and User ID reports) is excluded by default: the insights and the agent cannot read it until a super user unchecks Exclude visitor-level data.
• Report labels can hold values set by visitors (page titles, URLs, referrers, campaign names, custom dimensions). Check the retention and privacy terms of the provider before connecting it, and prefer a self-hosted or EU-hosted endpoint when your data policy requires it.
• The chat shows every user a notice stating that their questions, and the Matomo data read to answer them, are sent to the AI provider configured by the Matomo administrator.
• When a Matomo tool fails in agent mode, the model only receives a generic error with a reference: the details stay in the Matomo logs.
• Insight requests are restricted to the report and data methods of widgets declared by Matomo, never to an arbitrary API method, and run with the permissions of the current user.
• API keys are stored in the Matomo settings and are never sent back to the browser. Conversations are not stored by the plugin.

Need help with Matomo?

Openmost is an official Matomo Implementation Partner. We connect Matomo to AI assistants, BI tools and the rest of your stack with Matomo integrations built on official APIs, with documented and privacy-checked data flows.

• Agent mode weekly KPI review with tool steps

  

• Insights on any report in one click

  

• Insights panel follow up question with live data

  

• Settings connection and prompts

  

Features

Insights on any report

• An Insights button in the report header opens an insight panel that analyses the report you are looking at, then lets you ask follow-up questions.
• The whole report is analysed, not only the rows visible on screen: the plugin sends a compact payload with the rows, the report totals and the names and units of the metrics.
• The analysis follows what you see: the active segment, the period and date, and the compared periods or segments.
• The analysis also follows the table as you left it: the search, the sort, the flattening and up to 5 expanded rows, whose subtables are added to the analysis.
• Works with data tables, evolution graphs, goals, ecommerce, custom dimensions, custom reports, row evolution and the other reports Matomo declares as widgets.
• When a widget has no report data, or the report cannot be loaded (for example a missing access), a clear message is displayed instead of sending the error to the model.
• The panel is an accessible dialog: keyboard navigation, focus kept inside the panel, Escape to close.

Chat

• A ChatGPT page in the main menu, with suggested questions to start a conversation and a "New conversation" button.
• Streaming answers, with an automatic fallback when the endpoint does not stream.
• Auto-scroll that follows the answer while it is written, stops when you scroll up, and a button to jump back to the latest message.
• Markdown answers with scrollable tables and code blocks, and buttons to copy an answer or a code block.
• Enter sends the message, Shift+Enter adds a new line.

Agent mode, with AI Providers and MCP Server

When a provider is connected in AI Providers and the McpServer plugin is enabled, the chat and the insight panel work as an agent:

• The agent queries your live reports, websites, goals, dimensions and segments by itself, for any period, to answer with real figures.
• A timeline shows each Matomo tool used for an answer and its status.
• Tools run inside Matomo with the permissions of the current user: no public URL, OAuth client or extra token is needed.
• Write actions (for example creating a goal, an annotation or a segment) are only possible when McpServer allows write methods, and only after the agent has described the exact change and you have confirmed it explicitly in the conversation. This rule is added outside the editable prompts, so it cannot be removed by a custom prompt.
• When something is missing, the chat recommends the next step, one at a time: activate AI Providers, connect a provider, install, activate or enable MCP Server, allow write methods. Super users get a direct link for each step, other users are asked to contact their Matomo administrator.

The agent mode is optional. Without AI Providers or McpServer, the plugin keeps working with its own settings. With AI Providers but without McpServer, the connected provider answers without tools.

Connection and models

• One key is enough, used in this order: the API key set for the website, then the provider connected in AI Providers, then the general API key of the plugin. Keys are never copied from one place to another.
• When a provider is connected in AI Providers, it takes over: the host, API key and model of the general settings are displayed read-only, and are used again if AI Providers is deactivated. A key set for a website still overrides AI Providers for that website (the agent mode is then not used for that website).
• Works with OpenAI (default) and any OpenAI-compatible endpoint served over HTTPS, such as Azure OpenAI or a self-hosted server. The API key is optional on a custom host.
• The general API key is only sent to the general host: a website using its own host never receives it.
• Preset models, including Latest recommended (the default, which follows the model recommended by each plugin release, currently GPT 6 Astra), or any custom model name.
• When a model is deprecated, retired or not available to your OpenAI account, the chat explains it and links to the settings instead of showing the raw API error.

Prompts

• Default prompts written for analytics: the chat answers as a senior analytics consultant (direct answer, key figures, prioritised recommendations, no invented figures), and insights follow a fixed structure (summary, key figures, notable patterns, recommendations).
• Default prompts of previous versions are upgraded automatically to the new defaults, in the language of each user. Custom prompts are kept.
• A prompt equal to the default is not stored, so future improvements of the defaults reach you. A Reset to default button restores the default prompts.

More

• Interface translated into 13 languages: English, Arabic, Chinese (Simplified and Traditional), Dutch, French, German, Italian, Japanese, Polish, Portuguese, Spanish and Swedish.
• Follows the Matomo light and dark themes.
• Rate limit of 30 AI requests per hour, per user and per website, configurable in config/config.ini.php.
• HTTP API: ChatGPT.getResponse, ChatGPT.getStreamingResponse, ChatGPT.getInsights, ChatGPT.getSiteSettings, ChatGPT.setSiteSettings and ChatGPT.setSystemSettings.

Requirements

• Matomo 6.x (>=6.0.0-b1,<7.0.0-b1)
• PHP 8.1 or higher
• One of: an OpenAI API key, an OpenAI-compatible HTTPS endpoint, or a provider connected in AI Providers
• Optional, for the agent mode: the AI Providers plugin (bundled with Matomo) with a connected provider, and the McpServer plugin from the Marketplace. Write actions also require write methods to be allowed in the McpServer settings (Raw Matomo API tool access).

Installation / Configuration

1. Install and activate ChatGPT from Administration > Platform > Marketplace.
2. As a super user, open Administration > System > ChatGPT:
   • Connection card: host (an HTTPS URL), API key and model. Each card is saved on its own. The saved key is never displayed, and a Delete key button removes it.
   • Privacy card: check Allow sending Matomo data to the AI provider, off by default. Until a super user checks it, nothing is sent to the provider. The card shows where the data goes, and masks e-mail and IP addresses, removes URL query strings and excludes visitor-level data, all three on by default.
   • Prompts card: chat and insight base prompts, with a Reset to default button.
   • If a provider is connected in Administration > System > AI Providers, the connection fields are read-only and AI Providers is used.
3. Optionally, override the settings for a website in Administration > Websites > ChatGPT (website admin access): host, API key, model and prompts, with a Delete key button and a Use the general prompts button. Empty fields use the general settings.
4. For the agent mode, connect a provider in AI Providers, then install, activate and enable McpServer (Administration > System > General settings > McpServer). The chat guides you through each missing step.

5. Optionally, as a server administrator, tune the plugin in a [ChatGPT] section of config/config.ini.php. Every key is optional:

   [ChatGPT]
   ; name shown in the top menu, on the chat page and in the insights panel, 40 characters max
   display_name = "Company AI"
   ; AI requests per user and per website within the window, 0 disables the limit
   rate_limit_requests = 30
   ; length of the window, in seconds
   rate_limit_window = 3600
   

Then open the ChatGPT page in the main menu, or click the Insights button in the header of a report.

Privacy and data

• Insights send the data of the report you are looking at (labels, metrics and totals), its period, segment and comparisons, and the conversation, to the configured endpoint: OpenAI by default, your custom host, or the provider connected in AI Providers.
• The chat sends your messages and the prompt. In agent mode, the results of the Matomo tools the agent calls are also sent to the provider.
• Nothing is sent to the AI provider until a super user checks Allow sending Matomo data to the AI provider in Administration > System > ChatGPT, Privacy card. It is off by default: until then, the chat and the insights tell users to ask a super user to allow it.
• Before sending, e-mail and IP addresses are replaced with [email] and [ip], and URL query strings are removed, in the report data and in the results of the Matomo tools. Both are on by default and can be turned off in the same Privacy card.
• Visitor-level data (Visits Log, visitor profiles, real-time and User ID reports) is excluded by default: the insights and the agent cannot read it until a super user unchecks Exclude visitor-level data.
• Insight requests are restricted to the report and data methods of widgets declared by Matomo, never to an arbitrary API method, and run with the permissions of the current user.
• API keys are stored in the Matomo settings and are never sent back to the browser. Conversations are not stored by the plugin.

How do I install and configure this plugin?

1. Install and activate ChatGPT from Administration > Platform > Marketplace.
2. As a super user, open Administration > System > ChatGPT and set the host, API key and model in the Connection card, or connect a provider in Administration > System > AI Providers. In the Privacy card, check Allow sending Matomo data to the AI provider: it is off by default.
3. Optionally, override the settings for a website in Administration > Websites > ChatGPT.

You can also download the plugin from GitHub, extract it to your plugins/ folder and activate it.

Can I change the rate limit or the name shown in the menu?

Yes, in a [ChatGPT] section of config/config.ini.php: display_name renames the plugin in the top menu, on the chat page and in the insights panel, for example when you use a self-hosted model. rate_limit_requests sets the AI requests per user and per website (30 by default, 0 disables the limit) and rate_limit_window the length of the window in seconds (3600 by default). These values are kept when the plugin is updated.

The chat says that data sharing is not allowed

No Matomo data leaves your instance without an explicit decision: a super user must check Allow sending Matomo data to the AI provider in Administration > System > ChatGPT, Privacy card. Other users are asked to contact a super user.

What do I need to make it work?

One of: an OpenAI API key (https://platform.openai.com/), an OpenAI-compatible endpoint served over HTTPS, or a provider connected in AI Providers. On a custom host, the API key is optional: without a key, no Authorization header is sent.

Which API key is used?

One key is enough. The plugin uses, in this order: the API key set for the website in Administration > Websites > ChatGPT, then the provider connected in Administration > System > AI Providers, then the API key of the general settings. Keys are never copied from one place to another.

When AI Providers is connected, it takes over: the host, API key and model of the general settings are displayed read-only, and are used again if AI Providers is deactivated. A key set for a website still overrides AI Providers for that website.

How do I remove a saved API key?

Use the Delete key button of the Connection card, on the general settings page or on the settings page of the website. Leaving the field empty keeps the saved key.

Can I use models other than OpenAI's?

Yes. Any OpenAI-compatible chat completions endpoint served over HTTPS works, such as Azure OpenAI or a self-hosted server. You can also connect any provider supported by AI Providers. Type the model name in the Model (Custom) field when it is not in the preset list.

Which models are in the preset list?

• Latest recommended (default): follows the model recommended by each plugin release, currently GPT 6 Astra
• GPT 6 Astra, GPT 6.1 Sol, GPT 6 Sol, GPT 6 Luna
• GPT 5.6 Sol, GPT 5.6 Terra, GPT 5.6 Luna
• GPT 5.5, GPT 5.4, GPT 5.4 mini, GPT 5.4 nano
• GPT 5.2, GPT 5.1, GPT 5, GPT 5 mini, GPT 5 nano
• GPT 4.1, GPT 4.1 mini, GPT 4o, GPT 4o mini

Reasoning models (o-series), *-pro and *-codex variants are not listed: they are tuned for one-shot analysis or code, not for a conversation about report data.

The chat says my model is not available

OpenAI retires old models, and some models are not available to every account. Choose another model in the settings, for example "Latest recommended". The chat links to the settings page you can change.

What is the agent mode?

When a provider is connected in AI Providers and the McpServer plugin is enabled, the chat and the insight panel work as an agent: they query your live reports, websites, goals, dimensions and segments through the Matomo tools to answer with real figures. A timeline shows each tool used.

To enable it, connect a provider in AI Providers, then install, activate and enable McpServer in Administration > System > General settings > McpServer. The chat recommends each missing step, with a direct link for super users.

Can the agent change things in Matomo?

Only if McpServer allows write methods (Raw Matomo API tool access), and only with your confirmation: before any create, update or delete, the agent describes the exact change and waits for your explicit confirmation in the conversation. This rule cannot be removed by a custom prompt. The agent never has more access than the current user.

Do I need to expose my Matomo instance or configure OAuth for the agent?

No. The agent calls the Matomo tools inside your Matomo, with the permissions of the logged in user. It also works on private and intranet instances.

Does the plugin work without the agent mode?

Yes. Without AI Providers or McpServer, the chat and the insights use the host, API key and model of the plugin settings. With AI Providers but without McpServer, the connected provider answers without tools.

What do insights analyse?

The whole report you are looking at, not only the visible rows: its rows, totals and metrics, with the active segment, period and comparisons. Insights work on data tables, evolution graphs, goals, custom reports and the other reports Matomo declares as widgets.

Can I customise the prompts?

Yes, in the Prompts card of the general settings or of a website. The Reset to default button restores the default prompts, and on a website, Use the general prompts makes it follow the general prompts again. Default prompts of previous versions are upgraded automatically, custom prompts are kept.

Is my data sent to OpenAI?

Insights send the data of the report you are looking at (labels, metrics, totals, period, segment and comparisons) and the conversation to the configured endpoint: OpenAI by default, your custom host, or the provider connected in AI Providers. In agent mode, the results of the tools the agent calls are sent too. Raw visitor data is never sent, unless the report itself contains it, such as the Visits Log. Conversations are not stored by the plugin.

Is there a usage limit?

Yes, 30 AI requests per hour, per user and per website, to protect your API budget.

Which languages are supported?

English, Arabic, Chinese (Simplified and Traditional), Dutch, French, German, Italian, Japanese, Polish, Portuguese, Spanish and Swedish.

What are the requirements?

• Matomo 6.x
• PHP 8.1 or higher
• For the agent mode: a provider connected in AI Providers (bundled with Matomo) and the McpServer plugin

How do I get support?

Email ronan@openmost.com, or open an issue on GitHub. More on https://openmost.com/matomo/extensions/chatgpt.