Emotion Server API

Try Emotion Detection

Enter your own text or select from sample paragraphs to test our emotion detection API:

Key Features

The Emotion Server API is designed for developers and businesses seeking reliable emotion detection:

• 🧠 Fine-tuned Gemma3 model for emotion detection
• 🐳 Self-contained Docker solution for easy deployment
• 📁 GitHub repository with UI and MCP Server integration
• 📊 Cross-platform support (Mac Apple Silicon, Linux PC, Raspberry Pi)
• ✅ Fast, secure sentiment analysis with no external API calls
• 💡 Integration-ready for agentic solutions

Supported Emotions

The Emotion Server API detects text sentiment across 14 categories, each represented with its emoji:

API Usage

Deploy via Docker:

# For Raspberry Pi (32-bit ARM)
docker pull stevef1uk/emotion-emotion-service:arm

# For Apple Silicon (M1/M2/M3)
docker pull stevef1uk/emotion-service:arm64

# For standard x86_64 (Linux/Windows)
docker pull stevef1uk/emotion-service:amd64

# Run the container
docker run -d -p 8000:8000 stevef1uk/emotion-service:arm64

Send a POST request (fast path):

curl -s -X POST http://localhost:8000/predict \
     -H "Content-Type: application/json" \
     -d '{"text": "I’m so happy to see you!"}'

Response (fastest – confidence fixed at 1.0):
{
  "text": "I’m so happy to see you!",
  "emotion": "happiness",
  "confidence": 1,
  "emoji": "😊"
}

Optional: request accurate confidence (slower):

curl -s -X POST 'http://localhost:8000/predict?accurate=1' \
  -H 'Content-Type: application/json' \
  -d '{"text":"I am so happy today!"}' | jq .

# Default (fast path): /predict → confidence fixed at 1.0 (fastest)
# Accurate (slower): /predict?accurate=1 → confidence is the model's real probability

Detailed endpoint (public URL):

curl -s -X POST https://stevef1uk--emotion-server-serve.modal.run/predict_detailed \
     -H "Content-Type: application/json" \
     -d '{"text":"Im indifferent to the whole thing"}'

Response:
{
  "predicted_emotion": "neutral",
  "confidence": 0.75,
  "all_emotions": {
    "anger": 0.00,
    "confusion": 0.15,
    "desire": 0.00,
    "disgust": 0.06,
    "fear": 0.00,
    "guilt": 0.00,
    "happiness": 0.01,
    "love": 0.00,
    "neutral": 0.75,
    "sadness": 0.02,
    "sarcasm": 0.00,
    "shame": 0.01,
    "surprise": 0.00
  }
}

For a complete Docker Compose setup with a Gradio UI and MCP server as well as a tool to test the service against your input files, see the companion repo: emotion-service-demo on GitHub.

Demo Application: Watch the video below to see the Emotion Service API in action with a real-time dashboard showing emotion analysis of synthetic social media data:

Performance Benchmarks

Accuracy and speed compared across models (226 samples):

These results highlight Emotion Server API’s balance of accuracy and efficiency compared to alternatives.

Business Case

The Emotion Server API delivers secure, fast, and cost-effective emotion detection for enterprises and startups. Unlike cloud-only solutions, it runs locally within your environment, ensuring sensitive data is never exposed externally. With licensing options for indie developers, startups, and large enterprises, Emotion Server API is flexible and ready to scale with your needs.

Licensing Options

Note: All prices are subject to French TVA at the applicable rate (currently 20%).

📌 For licensing questions or bulk purchases, please contact: [email protected]
French VAT Number: FR05902395201

Frequently Asked Questions

• Does Emotion Server API require an internet connection?

  No, it runs entirely locally with Docker.

• Can I use it for commercial purposes?

  Yes, licensing options are available for startups and enterprises.

• Which platforms are supported?

  Mac (Apple Silicon), Linux, Windows (via Docker), and Raspberry Pi.

• Can I integrate this into existing software?

  Absolutely — API endpoints are standard RESTful JSON.

• Do you offer an evaluation license?

  Yes the docker container available today are licensed until the end of 2025

Contact & Support

For licensing inquiries, support, or technical assistance, please email [email protected] more details on GitHub.

Follow us on Social Media

Stay updated with the latest news and updates:

@ApiEmotion on X