Skip to content

davidmigloz/langchain_dart

Repository files navigation

πŸ¦œοΈπŸ”— LangChain.dart

tests docs langchain MIT Code Assist AI

Build LLM-powered Dart/Flutter applications.

What is LangChain.dart?

LangChain.dart is an unofficial Dart port of the popular LangChain Python framework created by Harrison Chase.

LangChain provides a set of ready-to-use components for working with language models and a standard interface for chaining them together to formulate more advanced use cases (e.g. chatbots, Q&A with RAG, agents, summarization, translation, extraction, recsys, etc.).

The components can be grouped into a few core modules:

LangChain.dart

  • πŸ“ƒ Model I/O: LangChain offers a unified API for interacting with various LLM providers (e.g. OpenAI, Google, Mistral, Ollama, etc.), allowing developers to switch between them with ease. Additionally, it provides tools for managing model inputs (prompt templates and example selectors) and parsing the resulting model outputs (output parsers).
  • πŸ“š Retrieval: assists in loading user data (via document loaders), transforming it (with text splitters), extracting its meaning (using embedding models), storing (in vector stores) and retrieving it (through retrievers) so that it can be used to ground the model's responses (i.e. Retrieval-Augmented Generation or RAG).
  • πŸ€– Agents: "bots" that leverage LLMs to make informed decisions about which available tools (such as web search, calculators, database lookup, etc.) to use to accomplish the designated task.

The different components can be composed together using the LangChain Expression Language (LCEL).

Motivation

Large Language Models (LLMs) have revolutionized Natural Language Processing (NLP), serving as essential components in a wide range of applications, such as question-answering, summarization, translation, and text generation.

The adoption of LLMs is creating a new tech stack in its wake. However, emerging libraries and tools are predominantly being developed for the Python and JavaScript ecosystems. As a result, the number of applications leveraging LLMs in these ecosystems has grown exponentially.

In contrast, the Dart / Flutter ecosystem has not experienced similar growth, which can likely be attributed to the scarcity of Dart and Flutter libraries that streamline the complexities associated with working with LLMs.

LangChain.dart aims to fill this gap by abstracting the intricacies of working with LLMs in Dart and Flutter, enabling developers to harness their combined potential effectively.

Packages

LangChain.dart has a modular design that allows developers to import only the components they need. The ecosystem consists of several packages:

Contains only the core abstractions as well as LangChain Expression Language as a way to compose them together.

Depend on this package to build frameworks on top of LangChain.dart or to interoperate with it.

Contains higher-level and use-case specific chains, agents, and retrieval algorithms that are at the core of the application's cognitive architecture.

Depend on this package to build LLM applications with LangChain.dart.

This package exposes langchain_core so you don't need to depend on it explicitly.

Contains third-party integrations and community-contributed components that are not part of the core LangChain.dart API.

Depend on this package if you want to use any of the integrations or components it provides.

Integration-specific packages

Popular third-party integrations (e.g. langchain_openai, langchain_google, langchain_ollama, etc.) are moved to their own packages so that they can be imported independently without depending on the entire langchain_community package.

Depend on an integration-specific package if you want to use the specific integration.

Package Version Description
langchain_anthropic langchain_anthropic Anthopic integration (Claude 3.5 Sonnet, Opus, Haiku, Instant, etc.)
langchain_chroma langchain_chroma Chroma vector database integration
langchain_firebase langchain_firebase Firebase integration (VertexAI for Firebase (Gemini 1.5 Pro, Gemini 1.5 Flash, etc.))
langchain_google langchain_google Google integration (GoogleAI, VertexAI, Gemini, PaLM 2, Embeddings, Vector Search, etc.)
langchain_mistralai langchain_mistralai Mistral AI integration (Mistral-7B, Mixtral 8x7B, Mixtral 8x22B, Mistral Small, Mistral Large, embeddings, etc.).
langchain_ollama langchain_ollama Ollama integration (Llama 3.2, Gemma 2, Phi-3.5, Mistral nemo, WizardLM-2, CodeGemma, Command R, LLaVA, DBRX, Qwen, Dolphin, DeepSeek Coder, Vicuna, Orca, etc.)
langchain_openai langchain_openai OpenAI integration (GPT-4o, o1, Embeddings, Tools, Vision, DALLΒ·E 3, etc.) and OpenAI Compatible services (TogetherAI, Anyscale, OpenRouter, One API, Groq, Llamafile, GPT4All, etc.)
langchain_pinecone langchain_pinecone Pinecone vector database integration
langchain_supabase langchain_supabase Supabase Vector database integration

API clients packages

The following packages are maintained (and used internally) by LangChain.dart, although they can also be used independently:

Depend on an API client package if you just want to consume the API of a specific provider directly without using LangChain.dart abstractions.

Package Version Description
anthropic_sdk_dart anthropic_sdk_dart Anthropic API client
chromadb chromadb Chroma DB API client
googleai_dart googleai_dart Google AI for Developers API client
mistralai_dart mistralai_dart Mistral AI API client
ollama_dart ollama_dart Ollama API client
openai_dart openai_dart OpenAI API client
openai_realtime_dart openai_realtime_dart OpenAI Realtime API client
tavily_dart tavily_dart Tavily API client
vertex_ai vertex_ai GCP Vertex AI API client

Integrations

The following integrations are available in LangChain.dart:

Chat Models

Chat model Package Streaming Multi-modal Tool-call Description
ChatAnthropic langchain_anthropic βœ” βœ” βœ” Anthropic Messages API (aka Claude API)
ChatFirebaseVertexAI langchain_firebase βœ” βœ” βœ” Vertex AI for Firebase API (aka Gemini API)
ChatGoogleGenerativeAI langchain_google βœ” βœ” βœ” Google AI for Developers API (aka Gemini API)
ChatMistralAI langchain_mistralai βœ” Mistral Chat API
ChatOllama langchain_ollama βœ” βœ” βœ” Ollama Chat API
ChatOpenAI langchain_openai βœ” βœ” βœ” OpenAI Chat API and OpenAI Chat API compatible services (GitHub Models, TogetherAI, Anyscale, OpenRouter, One API, Groq, Llamafile, GPT4All, FastChat, etc.)
ChatVertexAI langchain_google GCP Vertex AI Chat API

LLMs

Note: Prefer using Chat Models over LLMs as many providers have deprecated them.

LLM Package Streaming Description
Ollama langchain_ollama βœ” Ollama Completions API
OpenAI langchain_openai βœ” OpenAI Completions API
VertexAI langchain_google GCP Vertex AI Text API

Embedding Models

Embedding model Package Description
GoogleGenerativeAIEmbeddings langchain_google Google AI Embeddings API
MistralAIEmbeddings langchain_mistralai Mistral Embeddings API
OllamaEmbeddings langchain_ollama Ollama Embeddings API
OpenAIEmbeddings langchain_openai OpenAI Embeddings API
VertexAIEmbeddings langchain_google GCP Vertex AI Embeddings API

Vector Stores

Vector store Package Description
Chroma langchain_chroma Chroma integration
MemoryVectorStore langchain In-memory vector store for prototype and testing
ObjectBoxVectorStore langchain_community ObjectBox integration
Pinecone langchain_pinecone Pinecone integration
Supabase langchain_supabase Supabase Vector integration
VertexAIMatchingEngine langchain_google Vertex AI Vector Search (former Matching Engine) integration

Tools

Tool Package Description
CalculatorTool langchain_community To calculate math expressions
OpenAIDallETool langchain_openai OpenAI's DALL-E Image Generator
TavilyAnswerTool langchain_community Returns an answer for a query using the Tavily search engine
TavilySearchResultsTool langchain_community Returns a list of results for a query using the Tavily search engine

Getting started

To start using LangChain.dart, add langchain as a dependency to your pubspec.yaml file. Also, include the dependencies for the specific integrations you want to use (e.g.langchain_community, langchain_openai, langchain_google, etc.):

dependencies:
  langchain: {version}
  langchain_community: {version}
  langchain_openai: {version}
  langchain_google: {version}
  ...

The most basic building block of LangChain.dart is calling an LLM on some prompt. LangChain.dart provides a unified interface for calling different LLMs. For example, we can use ChatGoogleGenerativeAI to call Google's Gemini model:

final model = ChatGoogleGenerativeAI(apiKey: googleApiKey);
final prompt = PromptValue.string('Hello world!');
final result = await model.invoke(prompt);
// Hello everyone! I'm new here and excited to be part of this community.

But the power of LangChain.dart comes from chaining together multiple components to implement complex use cases. For example, a RAG (Retrieval-Augmented Generation) pipeline that would accept a user query, retrieve relevant documents from a vector store, format them using prompt templates, invoke the model, and parse the output:

// 1. Create a vector store and add documents to it
final vectorStore = MemoryVectorStore(
  embeddings: OpenAIEmbeddings(apiKey: openaiApiKey),
);
await vectorStore.addDocuments(
  documents: [
    Document(pageContent: 'LangChain was created by Harrison'),
    Document(pageContent: 'David ported LangChain to Dart in LangChain.dart'),
  ],
);

// 2. Define the retrieval chain
final retriever = vectorStore.asRetriever();
final setupAndRetrieval = Runnable.fromMap<String>({
  'context': retriever.pipe(
    Runnable.mapInput((docs) => docs.map((d) => d.pageContent).join('\n')),
  ),
  'question': Runnable.passthrough(),
});

// 3. Construct a RAG prompt template
final promptTemplate = ChatPromptTemplate.fromTemplates([
  (ChatMessageType.system, 'Answer the question based on only the following context:\n{context}'),
  (ChatMessageType.human, '{question}'),
]);

// 4. Define the final chain
final model = ChatOpenAI(apiKey: openaiApiKey);
const outputParser = StringOutputParser<ChatResult>();
final chain = setupAndRetrieval
    .pipe(promptTemplate)
    .pipe(model)
    .pipe(outputParser);

// 5. Run the pipeline
final res = await chain.invoke('Who created LangChain.dart?');
print(res);
// David created LangChain.dart

Documentation

Community

Stay up-to-date on the latest news and updates on the field, have great discussions, and get help in the official LangChain.dart Discord server.

LangChain.dart Discord server

Contribute

πŸ“’ Call for Collaborators πŸ“’
We are looking for collaborators to join the core group of maintainers.

New contributors welcome! Check out our Contributors Guide for help getting started.

Join us on Discord to meet other maintainers. We'll help you get your first contribution in no time!

Related projects

Sponsors

License

LangChain.dart is licensed under the MIT License.