Groq

Installation

<dependency>
    <groupId>com.google.genkit</groupId>
    <artifactId>genkit-plugin-groq</artifactId>
    <version>1.0.0-SNAPSHOT</version>
</dependency>

Configuration

export GROQ_API_KEY=your-api-key

Usage

import com.google.genkit.plugins.groq.GroqPlugin;

Genkit genkit = Genkit.builder()
    .plugin(GroqPlugin.create())
    .build();

ModelResponse response = genkit.generate(
    GenerateOptions.builder()
        .model("groq/llama-3.3-70b-versatile")
        .prompt("Tell me about AI")
        .build());

Available models

Model	Speed
groq/llama-3.1-8b-instant	~1200 tokens/sec
groq/llama-3.3-70b-versatile	~560 tokens/sec
groq/openai/gpt-oss-120b, groq/openai/gpt-oss-20b	Varies
groq/compound, groq/compound-mini	Agentic systems (web search + code execution)

Features

• Ultra-fast inference, text generation, streaming, tool calling, RAG, content moderation

Sample

See the groq sample.