Skip to content
The AI assistant for TIA Portal
TIA Portal AI assistant

The AI assistant for TIA Portal

Project-wide context · Agentic tools · Claude · GPT · Gemini · local · V15-V21 · and more
Not a chat next to TIA Portal, but an assistant that reads your project, generates code, tests it and imports with confirmation. With the LLM of your choice.

What the assistant can do

01

Understand the project

The assistant reads blocks, tags and references straight from the TIA project: explain unfamiliar logic, document legacy plants.
  • SCL, STL & graphical sources
  • Explain & document
  • Project-wide search
02

Generate code

A description becomes an SCL block using your project's real tag names and data types.
  • SCL generation
  • Convert STL to SCL
  • Refactoring
03

Work agentically

The assistant drives the whole Studio: export, import, edit, write unit tests, use EPLAN context, deploy Forge.
  • Full tool-set
  • Sub-agent delegation
  • Best-of-N in parallel
04

With confirmation

Nothing enters the project silently: diff preview, your approval, then the import over the official Openness API.
  • Diff before import
  • Openness API
  • No silent overwrites
05

Your model

Sign in or bring your own API key: Claude, GPT, Gemini, xAI and 20+ cloud providers. Or fully local via Ollama & co.
  • Multi-provider
  • Local via Ollama
  • No lock-in
06

Verify

Guard generated code immediately: SCL unit tests against PLCSIM Advanced or a real S7 PLC, and on request the AI writes the tests too.
  • PLCSIM Advanced
  • Real S7 PLC
  • AI-authored tests

TIA Portal AI assistant: how it works

A TIA Portal AI assistant is more than a chat window next to the engineering tool. The difference is access: the assistant in AnyAutomation Studio is connected to the project over the official Siemens Openness API (TIA Portal V15 through V21), knows blocks, tags and references, and can act on its own, from reading through generating to importing with your confirmation. The overview of the whole AI toolbox is on the AI for TIA Portal page.

Three steps

First: connect. Studio opens your TIA project over Openness. The assistant sees the project structure without you copying anything.

Second: assign. You describe the task in the chat ("explain FBConveyor", "generate an FB for the pallet turner with interlocking", "write unit tests for FBDosing"). The assistant uses its tools: read blocks, produce code, generate SCL, create tests.

Third: review and adopt. Every change arrives as a diff preview. You read, adjust and confirm, and only then does Studio import over Openness into the project. Afterwards, unit tests guard the result against PLCSIM Advanced or the real controller.

Agent, not chat

The assistant can delegate sub-tasks to specialised sub-agents and, with Best-of-N, run the same assignment in parallel across several isolated sessions: you compare the variants and adopt the best. Via the MCP client you connect your own tools, such as internal systems or check scripts. And because the assistant lives inside a complete engineering IDE, the results interlock: generated code gets versioned (Git), tested and documented instead of ending up in a chat history.

Your provider, your data

The assistant is multi-provider: Claude, GPT, Gemini, xAI via login or your own API key, plus 20+ cloud providers and enterprise clouds (Azure OpenAI, Vertex AI, Bedrock). For confidential plants, inference runs fully local via Ollama, LM Studio, vLLM or SGLang. Requests never pass through our server. How this differs from Siemens' own tooling is shown in the comparison on the Siemens Industrial Copilot alternative page.

Frequently asked questions

What is a TIA Portal AI assistant?

An AI assistant connected to the TIA project over the Openness API: it reads blocks and tags, generates and refactors SCL, writes unit tests and imports changes into the project after confirmation.

Which TIA Portal versions are supported?

TIA Portal V15 through V21 over the official Siemens Openness API, which includes legacy projects too old for Siemens' own AI tooling.

Which AI models can I use?

Claude, GPT, Gemini and Grok via login or API key, 20+ cloud providers (OpenRouter, Mistral, Groq, DeepSeek and more), enterprise clouds, and local servers (Ollama, LM Studio, vLLM, SGLang).

Does the assistant change my project without asking?

No. Every change arrives as a diff preview and only enters the project over Openness after your confirmation.

Does the assistant run offline?

Yes. With a local model via Ollama & co., inference runs entirely on your hardware, without the cloud.

What does the AI assistant cost?

The AI chat is included from the Basic plan (CHF 10 per month). Add your own API key with the model provider, or a local model with no key. Try free for 30 days.

Can the assistant write unit tests?

Yes. It produces SCL unit-test suites for the block and runs them against PLCSIM Advanced or a real S7 PLC. Details at SCL unit tests.

Is this a Siemens product?

No. AnyAutomation Studio is independently developed and uses the official Siemens Openness API. Siemens' own tools are Industrial Copilot and the Eigen Engineering Agent; the comparison shows the differences.

Try the assistant

Tell us your provider → we'll show the way.