cohere

Genkit Cohere Sample

This sample demonstrates integration with Cohere models using Genkit Java.

Features Demonstrated

• Cohere Plugin Setup - Configure Genkit with Command models
• Command A - Most capable and performant model (256K context)
• Command R - Balanced performance for content generation (128K context)
• Command R7B - Small, fast model for efficient tasks (128K context)
• Tool Usage - Search tool with automatic execution
• Streaming - Real-time response streaming

Supported Models

• command-a-03-2025 - Most capable Cohere model with 256K context and 150% higher throughput than Command R+ 08-2024
• command-r7b-12-2024 - Small and fast 7B model, 128K context
• command-r-08-2024 - Balanced performance model, 128K context
• command-r-plus-08-2024 - Previous flagship model, 128K context

Prerequisites

• Java 21+
• Maven 3.6+
• Cohere API key (get one at https://dashboard.cohere.com/)

Running the Sample

Option 1: Direct Run

# Set your Cohere API key
export COHERE_API_KEY=your-api-key-here

# Navigate to the sample directory
cd samples/cohere

# Run the sample
./run.sh
# Or: mvn compile exec:java

Option 2: With Genkit Dev UI (Recommended)

# Set your Cohere API key
export COHERE_API_KEY=your-api-key-here

# Navigate to the sample directory
cd samples/cohere

# Run with Genkit CLI
genkit start -- ./run.sh

The Dev UI will be available at http://localhost:4000

Available Flows

Flow	Model	Description
greeting	-	Simple greeting flow
chat	command-a-03-2025	Chat with Cohere's most capable model
researchAssistant	command-a-03-2025	Research assistant with search tool
streamingChat	command-a-03-2025	Streaming chat responses
generateContent	command-r-08-2024	Content generation
summarize	command-r7b-12-2024	Text summarization with small model
qa	command-r-08-2024	Question answering
creativeWriting	command-a-03-2025	Creative writing with streaming

Example API Calls

Once the server is running on port 8080:

Simple Greeting

curl -X POST http://localhost:8080/api/flows/greeting \
  -H 'Content-Type: application/json' \
  -d '"World"'

Chat

curl -X POST http://localhost:8080/api/flows/chat \
  -H 'Content-Type: application/json' \
  -d '"What are the benefits of AI?"'

Research Assistant (with tools)

curl -X POST http://localhost:8080/api/flows/researchAssistant \
  -H 'Content-Type: application/json' \
  -d '"Find information about renewable energy"'

Question Answering

curl -X POST http://localhost:8080/api/flows/qa \
  -H 'Content-Type: application/json' \
  -d '"What is machine learning?"'

Summarize

curl -X POST http://localhost:8080/api/flows/summarize \
  -H 'Content-Type: application/json' \
  -d '"[Your long text here]"'

Learn More

• Cohere Documentation
• Genkit Documentation