Skip to main content

​
Overview

FastEmbed provides fast, local embedding models powered by ONNX runtime. Supports GPU acceleration, sparse embeddings, and multiple model architectures including BGE, E5, and multilingual models. No API costs - runs entirely locally. Provider Class: FastEmbedProvider Config Class: FastEmbedConfig

​
Dependencies

uv pip install fastembed

​
Examples

from upsonic import Agent, Task, KnowledgeBase
from upsonic.embeddings import FastEmbedProvider, FastEmbedConfig
from upsonic.vectordb import ChromaProvider, ChromaConfig, ConnectionConfig, Mode

# Create embedding provider
embedding = FastEmbedProvider(FastEmbedConfig(
    model_name="BAAI/bge-small-en-v1.5",
    enable_gpu=True
))

# Setup KnowledgeBase
vectordb = ChromaProvider(ChromaConfig(
    collection_name="fastembed_docs",
    vector_size=384,
    connection=ConnectionConfig(mode=Mode.IN_MEMORY)
))

kb = KnowledgeBase(
    sources=["document.txt"],
    embedding_provider=embedding,
    vectordb=vectordb
)

# Query with Agent
agent = Agent("anthropic/claude-sonnet-4-5")
task = Task("What is this document about?", context=[kb])
result = agent.do(task)
print(result)

​
Parameters

ParameterTypeDescriptionDefaultSource
model_namestrFastEmbed model name"BAAI/bge-small-en-v1.5"Specific
cache_dirstr | NoneModel cache directoryNoneSpecific
threadsint | NoneNumber of threads (auto-detected if None)NoneSpecific
providerslist[str]ONNX execution providers["CPUExecutionProvider"]Specific
enable_gpuboolEnable GPU acceleration if availableFalseSpecific
enable_parallel_processingboolEnable parallel text processingTrueSpecific
doc_embed_typestrDocument embedding type (default, passage)"default"Specific
max_memory_mbint | NoneMaximum memory usage in MBNoneSpecific
model_warmupboolWarm up model on initializationTrueSpecific
enable_sparse_embeddingsboolUse sparse embeddings for better performanceFalseSpecific
sparse_model_namestr | NoneSparse model name if different from denseNoneSpecific
batch_sizeintBatch size for document embedding100Base
max_retriesintMaximum number of retries on failure3Base
normalize_embeddingsboolWhether to normalize embeddings to unit lengthTrueBase
show_progressboolWhether to show progress during batch operationsTrueBase