🧠 LFM2.5 Embedding API

OpenAI-compatible embeddings running on pure CPU

Operational

📊 Model Information

Model

LFM2.5-Embedding-350M

Provider

Liquid AI

Dimensions

1024 floats

Context

512 tokens

Quantization

Q8_0 (GGUF)

Engine

llama.cpp (CPU)

🔌 Endpoint

POST/v1/embeddings

Compatible with OpenAI SDK, LangChain, LlamaIndex, Open WebUI, Cursor, OpenClaw, and more.

🚀 Quick Start (cURL)

curl -X POST "https://huggingface.co/proxy/YOUR-SPACE.hf.space/v1/embeddings" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": "Your text here",
    "model": "LiquidAI/LFM2.5-Embedding-350M",
    "input_type": "document"
  }'

🐍 Python Example

from openai import OpenAI

client = OpenAI(
    api_key="YOUR_API_KEY",
    base_url="https://huggingface.co/proxy/YOUR-SPACE.hf.space/v1/"
)

response = client.embeddings.create(
    model="LiquidAI/LFM2.5-Embedding-350M",    input="Your text here",
    extra_body={"input_type": "document"}
)

print(response.data[0].embedding[:5])

🔐 Authentication

All requests require a Bearer token:

Authorization: Bearer YOUR_API_KEY

📚 Documentation

For complete setup instructions and licensing:

View README →