nlweb-models

Bundle package containing all LLM and embedding model providers for NLWeb.

Included Providers

LLM Providers

• OpenAI - GPT-3.5, GPT-4, and newer models
• Anthropic - Claude models
• Google Gemini - Gemini Pro and variants
• Azure OpenAI - Azure-hosted OpenAI models with managed identity support
• Azure Llama - Llama models via Azure
• Azure DeepSeek - DeepSeek models via Azure
• HuggingFace - Models from HuggingFace Hub
• Snowflake - Snowflake Cortex LLM
• Ollama - Local LLM serving
• Inception - Custom LLM provider

Embedding Providers

• OpenAI - text-embedding-ada-002, text-embedding-3-small/large
• Azure OpenAI - Azure-hosted OpenAI embeddings
• Google Gemini - Gemini embedding models
• Snowflake - Snowflake Cortex embeddings
• Ollama - Local embedding models
• Elasticsearch - Elasticsearch ML embeddings

Installation

pip install nlweb-core nlweb-models

Configuration Examples

OpenAI

llm:
  provider: openai
  import_path: nlweb_models.llm.openai
  class_name: provider
  api_key_env: OPENAI_API_KEY
  models:
    high: gpt-4
    low: gpt-3.5-turbo

embedding:
  provider: openai
  import_path: nlweb_models.embedding.openai_embedding
  class_name: get_openai_embeddings
  api_key_env: OPENAI_API_KEY
  model: text-embedding-3-small

Anthropic Claude

llm:
  provider: anthropic
  import_path: nlweb_models.llm.anthropic
  class_name: provider
  api_key_env: ANTHROPIC_API_KEY
  models:
    high: claude-3-opus-20240229
    low: claude-3-haiku-20240307

Azure OpenAI

llm:
  provider: azure_openai
  import_path: nlweb_models.llm.azure_oai
  class_name: provider
  endpoint_env: AZURE_OPENAI_ENDPOINT
  api_key_env: AZURE_OPENAI_KEY
  api_version: 2024-02-01
  auth_method: azure_ad  # or api_key
  models:
    high: gpt-4
    low: gpt-35-turbo

embedding:
  provider: azure_openai
  import_path: nlweb_models.embedding.azure_oai_embedding
  class_name: get_azure_embedding
  endpoint_env: AZURE_OPENAI_ENDPOINT
  auth_method: azure_ad
  model: text-embedding-ada-002

Google Gemini

llm:
  provider: gemini
  import_path: nlweb_models.llm.gemini
  class_name: provider
  api_key_env: GOOGLE_API_KEY
  models:
    high: gemini-pro
    low: gemini-pro

embedding:
  provider: gemini
  import_path: nlweb_models.embedding.gemini_embedding
  class_name: get_gemini_embeddings
  model: models/embedding-001

Usage

import nlweb_core

# Initialize with config
nlweb_core.init(config_path="./config.yaml")

# Use LLM
from nlweb_core import llm

result = await llm.ask_llm(
    prompt="Summarize this text",
    schema={"type": "object", "properties": {"summary": {"type": "string"}}},
    level="high"
)

# Use embeddings
from nlweb_core import embedding

vector = await embedding.get_embedding(
    text="Text to embed"
)

Provider Import Paths

LLM Providers

Embedding Providers

License

MIT License - Copyright (c) 2025 Microsoft Corporation