Ollama Model

agentscope-extensions-model-ollama integrates locally hosted Ollama models. It is useful for local development, private deployments, and offline model serving.

Add the dependency

<dependency>
    <groupId>io.agentscope</groupId>
    <artifactId>agentscope-extensions-model-ollama</artifactId>
    <version>${agentscope.version}</version>
</dependency>

ModelRegistry

Use the ollama:<model> id. OLLAMA_BASE_URL is optional and defaults to the local Ollama endpoint when omitted.

ReActAgent agent = ReActAgent.builder()
    .name("assistant")
    .model("ollama:llama3") // Resolved internally by ModelRegistry.resolve(modelId)
    .build();

Explicit builder

Use the builder when you need a non-default Ollama endpoint, formatter, transport, proxy, or Ollama options:

import io.agentscope.extensions.model.ollama.OllamaChatModel;

OllamaChatModel model = OllamaChatModel.builder()
    .modelName("llama3")
    .baseUrl("http://localhost:11434")
    .build();

Spring Boot

Spring Boot applications can use the Ollama starter:

<dependency>
    <groupId>io.agentscope</groupId>
    <artifactId>agentscope-ollama-spring-boot-starter</artifactId>
    <version>${agentscope.version}</version>
</dependency>

Configure the local Ollama model with agentscope.model.provider=ollama. The base URL is optional and defaults to http://localhost:11434:

agentscope:
  model:
    provider: ollama
  ollama:
    model-name: llama3
    # base-url: http://localhost:11434

Full builder options, formatters, credentials, and registry context details are covered in Model.