dot CMS

dotAI

September 24, 2026

dotAI

Description

dotAI is the AI engine built into dotCMS. It brings generative and semantic AI into the places teams already work (the Block Editor, workflows, search, and developer APIs), so teams can create content faster, automate repetitive tasks, and give visitors search that understands what they mean.

dotAI does three things, and each can be configured, tested, and switched on or off independently:

  • Chat (text generation) powers the AI Content block in the Block Editor, workflow actions that write copy, generate SEO metadata, translate, auto-tag content, and write image alt text in bulk, and the REST endpoints and $ai Velocity viewtool your own code calls.

  • Embeddings turn your content into a semantic index. Semantic search queries that index, and it grounds chat answers in your own content instead of the model's general knowledge.

  • Image generation creates images from a prompt, on demand or through a workflow action that writes the result straight into a content field.

dotAI runs on the AI provider you choose, from seven supported options: OpenAI, Azure OpenAI, Amazon Bedrock, and Google Vertex AI to run inside your own cloud; Anthropic and Google AI (Gemini) with a single API key; and OpenRouter, which reaches hundreds of models through one key. You can mix providers by capability, for example chat on one and embeddings on another.

Two admin screens tie it together:

  • dotAI App Configuration is where you connect a provider. Pick a provider per capability, fill in only the fields it requires, and run Test Connection to confirm your endpoint and credentials work before saving. Secrets stay on the server and are never shown in the browser.

  • The dotAI Tool (Developer Tools → dotAI) is where you try AI on your own content. Its five tabs let you run semantic search, chat with your content, generate images, build and manage embeddings indexes, and inspect every resolved setting in Config Values.

dotAI was first released as a core dotCMS feature in 24.04.05. Since then it has had three major updates:

  • Multi-Provider Support (26.04.25-01, all seven providers in 26.07.04-01): dotAI capabilities can run on any of seven providers instead of OpenAI only, with chat, embeddings, and image each set separately. Existing OpenAI setups keep working unchanged.

  • New dotAI Tool (26.09.09-01): rebuilt in the modern admin with five URL-addressable tabs, streaming chat answers you can stop mid-response, ranked search results with closeness scores, retrieval settings that persist between visits, guided index management, and your installation's theme and language.

  • dotAI App Configuration screen (26.09.24-01): a guided form replaces hand-written JSON, with per-capability setup, field validation, Test Connection, and read-only secrets. Automated setup through the providerConfig JSON contract still works.


Benefits

For Content and Marketing Teams

  • Draft and refine copy and generate images inside the editor, without switching to outside AI tools.

  • Automate SEO metadata, summaries, translations, tags, and alt text at publish time or across large batches of content.

  • Give visitors search and chat that understand plain-language questions and answer from your own content.

For Developers and Platform Teams

  • Build AI into workflows, templates, and headless apps with workflow actions, the $ai viewtool, and REST endpoints for generation, search, and embeddings.

  • Choose the provider and model in configuration, with no code changes. Run in Azure, AWS, or Google Cloud, route through an internal gateway, and use cheaper models for simpler jobs.

  • Prove a setup works with Test Connection, then use the dotAI Tool to tune semantic search against real content before you put it in front of users.

For Architects and Security Teams

  • Keep AI inside the boundary your organization already approves by pointing dotAI at your contracted cloud or an approved provider.

  • Credentials stay on the server and are redacted in logs, and the Config Values tab shows each resolved setting and where it came from.


Use Cases

AI-assisted content creation
A marketing team drafts blog posts and campaign copy with the AI Content block, refines tone before publishing, and generates supporting images from a prompt.

Automated publishing workflows
A publishing team adds workflow actions that generate SEO metadata, summaries, tags, and image alt text whenever content is published, with no manual step.

Conversational search on your own content
A documentation site replaces keyword search with dotAI semantic search and chat. A developer first builds an index in the dotAI Tool, runs the questions users actually ask, and adjusts the retrieval threshold until the results are right.

Clearing an AI security review
An architect at a financial-services company points dotAI at the company's Azure OpenAI or Amazon Bedrock account and runs Test Connection with the security team watching. Content teams then use the same AI writing, tagging, and search, now running inside the approved cloud.

Choosing the right model for each job
A platform team runs chat on one provider and embeddings on another, or uses OpenRouter to switch among hundreds of models by changing a setting, picking models for cost, speed, or where they are allowed to run.


Resources

Explore dotCMS for your organization

image

dotCMS Named a Major Player

In the IDC MarketScape: Worldwide AI-Enabled Headless CMS 2025 Vendor Assessment

image

Explore an interactive tour

See how dotCMS empowers technical and content teams at compliance-led organizations.

image

Built for Compliance. Certified for AI.

dotCMS is ISO 27001 and ISO 42001 certified, pairing independently verified information security with governed, accountable AI.