Telusko Notes

GeminiSDKDemo – Calling Gemini Model

• GeminiSDKDemo is a simple Java program that calls the Gemini model using the
com.google.genai.Client class.

• The code uses the model name "gemini-2.5-flash" and sends a text prompt asking for one
comedy movie with SRK, just the name.

• The response from the model is stored in a GenerateContentResponse object and printed using
response.text().

• Authentication for the Gemini client is provided via the GOOGLE_API_KEY environment
variable.

Code Implementation:

package com.telusko;

import com.google.genai.Client;
import com.google.genai.types.GenerateContentResponse;

public class GeminiSDKDemo {
    public static void main(String[] args) {
        Client client = Client.builder()
                .apiKey(System.getenv("GOOGLE_API_KEY"))
                .build();

        GenerateContentResponse response =
                client.models.generateContent(
                        "gemini-2.5-flash",
                        "name one comedy movie with SRK, just the name"
                );

        System.out.println(response.text());
    }
}

OllamaSDKDemo – Calling Local Mistral Model

• OllamaSDKDemo is a Java program that connects to a local Ollama server running at
"http://localhost:11434/".

• It creates an OllamaAPI client with the host URL and prepares a prompt asking for one comedy
movie with Shah Rukh Khan, just the name.

• The code calls client.generate("mistral", prompt, false, false, options) using an Options object
from OptionsBuilder.

• The result is printed using result.getResponse(), and any OllamaBaseException, IOException,
or InterruptedException is caught and logged with e.printStackTrace()

Code Implementation:


package com.telusko;

import io.github.ollama4j.OllamaAPI;
import io.github.ollama4j.exceptions.OllamaBaseException;
import io.github.ollama4j.models.response.OllamaResult;
import io.github.ollama4j.utils.Options;
import io.github.ollama4j.utils.OptionsBuilder;

import java.io.IOException;

public class OllamaSDKDemo {
    public static void main(String[] args) {

        String host = "http://localhost:11434/";
        OllamaAPI client = new OllamaAPI(host);

        String prompt = "name one comedy movie with Shah Rukh Khan, just the name";

        Options options = new OptionsBuilder().build();

        try {
            OllamaResult result = client.generate(
                    "mistral",
                    prompt,
                    false,
                    false,
                    options
            );

            System.out.println(result.getResponse());

        } catch (OllamaBaseException | IOException | InterruptedException e) {
            e.printStackTrace();
        }
    }
}

OpenAISDKDemo – Calling OpenAI gpt-5 Model

• OpenAISDKDemo is a Java program that uses OpenAIClient created via
OpenAIOkHttpClient.fromEnv() to talk to OpenAI.

• It builds ResponseCreateParams with an input prompt asking for one comedy movie with SRK,
just the name.

• The model name is set to "gpt-5" inside ResponseCreateParams.builder().model("gpt-5").

• The response is obtained using client.responses().create(params) and printed directly with
System.out.println(response).

Code Implementation:

package com.telusko;

import com.openai.client.OpenAIClient;
import com.openai.client.okhttp.OpenAIOkHttpClient;
import com.openai.models.responses.Response;
import com.openai.models.responses.ResponseCreateParams;

public class OpenAISDKDemo {
    public static void main(String[] args) {

        OpenAIClient client = OpenAIOkHttpClient.fromEnv();

        ResponseCreateParams params =
                ResponseCreateParams.builder()
                        .input("name one comedy movie with SRK, just the name")
                        .model("gpt-5")
                        .build();

        Response response = client.responses().create(params);

        System.out.println(response);
    }
}

Dependencies in pom.xml

<dependencies>

    <!-- Dependency for OpenAI -->
    <dependency>
        <groupId>com.openai</groupId>
        <artifactId>openai-java</artifactId>
        <version>4.0.0</version>
    </dependency>

    <!-- Dependency for Gemini -->
    <dependency>
        <groupId>com.google.genai</groupId>
        <artifactId>google-genai</artifactId>
        <version>1.0.0</version>
    </dependency>

    <!-- Dependency for Ollama -->
    <dependency>
        <groupId>io.github.ollama4j</groupId>
        <artifactId>ollama4j</artifactId>
        <version>1.1.0</version>
    </dependency>

</dependencies>

Common Pattern Across All.

• All three programs create a client object specific to the provider SDK: GeminiClient,
OllamaAPI, and OpenAIClient.

• In every case, the code sends the prompt to a model and prints the returned response to the
console.

• Configuration details such as environment variables (GOOGLE_API_KEY, OpenAI env) or
the localhost URL (http://localhost:11434/) are used to connect to each respective service.

Exploring SDKs through Java

GeminiSDKDemo – Calling Gemini Model

OllamaSDKDemo – Calling Local Mistral Model

OpenAISDKDemo – Calling OpenAI gpt-5 Model

Dependencies in pom.xml

Common Pattern Across All.

On this page