#Is it possible to turn a locally running model into an full agent in IntelliJ IDEA?

1 messages · Page 1 of 1 (latest)

crimson dagger
#

Hello, I am running a locally a model through LM Studio, but problem is that I can only connect it to chat IntellijIDEA. Is it possible to turn a locally running model into an full agent in IntelliJ IDEA? thanks for answer

normal hawkBOT
#

<@&987246527741304832> please have a look, thanks.

#

Here is an AI assisted attempt to answer your question 🤖. Maybe it helps! In any case, a human is on the way 👍. To continue talking to the AI, you can use </chatgpt:1108714622413963314>.

#
TJ-Bot
Hello, I am running a locally a model through LM Studio, but problem is that I can only connect it to chat IntellijIDEA. Is it possible to turn a locally running model into an full agent in IntelliJ IDEA? thanks for answer

Turning a locally running LLM (like the one managed by LM Studio) into a full agent within IntelliJ IDEA depends on the integration capabilities and plugins IntelliJ supports. Currently, IntelliJ IDEA lets you connect to local or remote LLMs primarily for code completion, chat, or assistance features through plugins or API hooks. However, transforming that model into a full agent — meaning one that can autonomously run complex workflows, handle tasks beyond chat (like code generation combined with execution, debugging, or project orchestration) — typically requires custom development. This could involve:

  • Building a plugin or extension that interfaces your local model with IntelliJ’s project and run configurations.
  • Using an orchestration layer that connects to your model and IntelliJ environment to send commands and receive actionable results.
  • Leveraging existing agent frameworks designed for code environments and adapting them to your local model.

In short, IntelliJ IDEA doesn't natively turn a chat-connected local model into a full autonomous agent, but with development, it’s feasible to build such integration. Many professional setups leverage middleware or custom tooling around LLMs to create agents that interact deeply with IDEs.

Useful links:

- AI generated response using gpt-4.1-mini model
dawn nebula
#

You can configure other model providers in the settings. They would need to implement one of the supported APIs - being compatible with OpenAI's API seems to be a popular choice.

I haven't done it myself though.

crimson dagger
#

i cannot find where can i set model provider for the agent. i did found some short tutorials where they are editing some JSOn files, but i cannot find where this json files are

crimson dagger
#

OK, it is possible via the ACP agent OpenCode. But this agent is configured in the opencode.json file. And you can't find information anywhere on the internet about where this file should be located. This is crazy. I found it randomly.