Enterprise Plans
AI Application Development
AI Translator API
AI SEO/GEO Service
Web Scraping Service
Log in
const { OpenAI } = require('openai');
const main = async () => {
const api = new OpenAI({ apiKey: '', baseURL: 'https://api.ai.cc/v1' });
const text = 'Your text string goes here';
const response = await api.embeddings.create({
input: text,
model: 'text-embedding-3-large',
});
const embedding = response.data[0].embedding;
console.log(embedding);
};
main();
import json
from openai import OpenAI
def main():
client = OpenAI(
base_url="https://api.ai.cc/v1",
api_key="",
)
text = "Your text string goes here"
response = client.embeddings.create(input=text, model="text-embedding-3-large")
embedding = response.data[0].embedding
print(json.dumps(embedding, indent=2))
main()
-
AI Playground
Test all API models in the sandbox environment before you integrate.
We provide more than 300 models to integrate into your app.
Product Detail
Introducing text-embedding-3-small: OpenAI's Latest Innovation in Text Embeddings
Released by OpenAI on January 25, 2024, text-embedding-3-small is a cutting-edge text embedding model engineered for superior performance and efficiency. This model marks a significant advancement, designed to transform diverse text inputs into compact, numerical representations (embeddings) that are highly effective for various machine learning applications. It serves as a powerful successor to text-embedding-ada-002, offering enhanced capabilities across the board.
🌟 Key Features & Advantages
- ✅ Enhanced Performance: Achieves remarkable improvements in multi-language retrieval (MIRACL) and English-specific tasks (MTEB), making it more robust and accurate.
- 💰 Cost Efficiency: Experience a substantial 5x reduction in cost compared to its predecessor, text-embedding-ada-002, offering significant savings for developers and businesses.
- 📏 Compact Size: With an embedding size of 512 dimensions, this model is ideal for environments with memory and storage constraints, ensuring efficient operation without compromising quality.
🚀 Versatile Applications
The text-embedding-3-small model is engineered for a wide array of applications, enabling intelligent text analysis and integration:
- Intelligent Search: Elevate search algorithms by accurately ranking results based on semantic relevance.
- Text Clustering: Group similar text documents or strings for advanced data analysis and organization.
- Recommendation Systems: Power sophisticated recommendation engines by suggesting related items based on text similarity.
- Anomaly Detection: Identify unusual patterns or outliers within large datasets with higher precision.
- Diversity Measurement: Analyze the breadth and variety of text data for deeper insights.
- Content Classification: Classify text strings by associating them with their most semantically similar labels.
🌐 Extensive Language Support
Designed for a global audience, text-embedding-3-small offers robust support for multiple languages, significantly enhancing its accessibility and utility across diverse linguistic datasets and international applications.
⚙️ Technical Specifications
- Architecture: The model leverages a state-of-the-art transformer-based architecture, meticulously optimized for both computational efficiency and high-performance embedding generation.
- Training Data: Trained on an extensive and diverse collection of text sources, ensuring it captures a broad spectrum of linguistic patterns and semantic nuances. This comprehensive training minimizes bias and ensures robust performance across varied demographics and use cases.
- Data Source and Size: Encompasses millions of text documents, providing the model with a profound understanding of language complexities and contexts.
📊 Performance Benchmarks
The text-embedding-3-small model sets new standards in embedding performance:
- ⭐ Significant Improvements Over text-embedding-ada-002:
- MIRACL Score: Increased from 31.4% to 44.0% (a notable 12.6% improvement).
- MTEB Score: Improved from 61.0% to 62.3%.
- 🎯 Higher Accuracy: Demonstrates superior accuracy across both multi-language and English-specific benchmarks, delivering more precise and reliable embeddings.
- ⚡ Enhanced Speed: Operates with greater efficiency compared to previous models, leading to reduced latency and lower computational resource requirements.
- 🛡️ Robustness: Capable of handling diverse input types effectively, ensuring consistent and dependable performance across a wide range of applications and data complexities.
❓ Frequently Asked Questions (FAQ)
Q1: What is text-embedding-3-small?
A1: It's OpenAI's latest text embedding model, released on January 25, 2024, designed to convert text into efficient numerical representations (embeddings) for machine learning tasks, offering enhanced performance and cost efficiency over its predecessors.
Q2: How does it compare to text-embedding-ada-002?
A2: text-embedding-3-small offers significantly improved performance (e.g., higher MIRACL and MTEB scores) and is 5x more cost-efficient than text-embedding-ada-002, while maintaining a compact embedding size.
Q3: What are the primary use cases for this model?
A3: It's ideal for a wide range of applications including intelligent search, text clustering, recommendation systems, anomaly detection, diversity measurement, and text classification across various languages.
Q4: Is text-embedding-3-small suitable for multi-language applications?
A4: Yes, it boasts extensive multi-language support and shows significant performance improvements on multi-language retrieval benchmarks (MIRACL), making it highly suitable for diverse linguistic datasets.
Q5: What is the embedding dimension of text-embedding-3-small?
A5: The model generates embeddings with a compact size of 512 dimensions, making it efficient for memory and storage-constrained environments.
Learn how you can transformyour company with AICC APIs
