catsu 0.1.4

High-performance embeddings client for multiple providers

Catsu

High-performance embeddings client for multiple providers, powered by Rust.

Installation

pip install catsu

Quick Start

from catsu import Client

# Create client (reads API keys from environment)
client = Client()

# Generate embeddings
response = client.embed(
    "openai:text-embedding-3-small",
    ["Hello, world!", "How are you?"]
)

print(f"Dimensions: {response.dimensions}")
print(f"Tokens used: {response.usage.tokens}")

Async Support

import asyncio
from catsu import Client

async def main():
    client = Client()
    response = await client.aembed("openai:text-embedding-3-small", "Hello!")
    print(response.embeddings[0][:5])

asyncio.run(main())

Model Catalog

# List all available models
models = client.list_models()

# Filter by provider
openai_models = client.list_models("openai")
for m in openai_models:
    print(f"{m.name}: {m.dimensions} dims, ${m.cost_per_million_tokens}/M tokens")

Configuration

client = Client(
    max_retries=5,   # Default: 3
    timeout=60,      # Default: 30 seconds
)

Supported Providers

• OpenAI (OPENAI_API_KEY)
• VoyageAI (VOYAGE_API_KEY)
• Cohere (COHERE_API_KEY)
• Jina (JINA_API_KEY)
• Mistral (MISTRAL_API_KEY)
• Gemini (GOOGLE_API_KEY or GEMINI_API_KEY)
• Together (TOGETHER_API_KEY)
• Mixedbread (MIXEDBREAD_API_KEY)
• Nomic (NOMIC_API_KEY)
• DeepInfra (DEEPINFRA_API_KEY)
• Cloudflare (CLOUDFLARE_API_TOKEN + CLOUDFLARE_ACCOUNT_ID)

Features

• 11 providers with unified API
• Async support with aembed()
• Model catalog with 64+ models
• Automatic retry with exponential backoff
• Cost tracking per request
• High performance - Rust core with Python bindings