All products
Production-ready toolkit for enterprise LangGraph agents

Fastal LangGraph Toolkit

Open source toolkit for enterprise LangGraph agents: multi-provider support, intelligent conversation management and production-ready speech processing.

Request information

Key Features

Multi-Provider Model Factory

Support for OpenAI (including GPT-5), Anthropic, Ollama, AWS Bedrock with seamless provider switching.

Intelligent Conversation Management

Summarization based on conversation structure, not just token counting.

Enterprise Speech Processing

Production-ready speech-to-text processing with multi-format support and language detection.

Memory Management

Type-safe SummarizableState with native LangGraph integration and token optimization.

A toolkit born from experience

The Fastal LangGraph Toolkit was originally developed internally by the Fastal Group to support enterprise-grade agentic application implementations across multiple client projects. After proving its effectiveness in production environments, we’ve open-sourced this toolkit to contribute to the broader LangGraph community.

pip install fastal-langgraph-toolkit

View package on PyPI

Why this toolkit

Building production LangGraph agents involves solving common challenges in advanced research and development projects:

  • Multi-provider Management — Support for multiple LLM/embedding/speech providers with seamless switching
  • Context Management — Intelligent conversation summarization for long-running sessions
  • Memory Optimization — Token-efficient context handling for cost control
  • Speech Processing — Enterprise-grade speech-to-text transcription capabilities
  • Type Safety — Proper state management with TypedDict integration
  • Configuration Injection — Clean separation between business logic and framework concerns

This toolkit provides battle-tested solutions for these challenges, extracted from real enterprise implementations.

Main features

Multi-Provider Model Factory

The current version of the model factory supports the following providers:

LLM Support: OpenAI (including GPT-5 models), Anthropic, Ollama, AWS Bedrock

Embeddings Support: OpenAI, Ollama, AWS Bedrock

Speech-to-Text Support: OpenAI Whisper (more providers coming soon)

Features:

  • Full GPT-5 Support: GPT-5, GPT-5-mini, and GPT-5-nano with automatic parameter mapping
  • Configuration Injection: Clean provider abstraction
  • Provider Health Checks: Availability validation
  • Seamless Switching: Change providers without code changes

Enterprise Speech Processing

Production-ready speech-to-text processing with enterprise-grade reliability and performance.

  • Multi-Format Support: MP3, WAV, M4A, and other common audio formats
  • Language Detection: Automatic language identification and custom language hints
  • Async Processing: Full async/await support for non-blocking operations
  • Segment Information: Detailed timestamp and confidence data when available
  • Error Handling: Robust error management with detailed logging
  • Lazy Loading: Efficient resource management with provider lazy loading

Intelligent Conversation Summarization

The LangChain/LangGraph framework provides good support for managing both short-term and long-term memory in agents through the LangMem module. However, we found that automated summarization based solely on token counting is not a sufficient approach for most real and complex agents.

The solution included in this kit offers an alternative and more sophisticated method, based on the structure of the conversation and a focus on the object and content of the discussions.

  • Ready-to-Use LangGraph Node: summary_node() method provides instant integration
  • Conversation Pair Counting: Smart Human+AI message pair detection
  • ReAct Tool Filtering: Automatic exclusion of tool calls from summaries
  • Configurable Thresholds: Customizable trigger points for summarization
  • Context Preservation: Keep recent conversations for continuity
  • Custom Prompts: Domain-specific summarization templates
  • State Auto-Injection: Seamless integration with existing states

Memory Management

  • SummarizableState: Type-safe base class for summary-enabled states
  • Automatic State Management: No manual field initialization required
  • LangGraph Integration: Native compatibility with LangGraph checkpointing
  • Clean Architecture: Separation of concerns between summary and business logic

Benefits

  • Open source Python package on PyPI
  • Battle-tested in real enterprise production projects
  • Full support for GPT-5, GPT-5-mini, GPT-5-nano
  • Native integration with LangGraph checkpointing

Interested?

Contact us for a demo or to receive more information.

Request demo

Frequently asked questions

Why did Fastal open source this toolkit?
It was built in-house to support enterprise implementations of agentic applications across several client projects. Having proven its effectiveness in real production environments, the Group chose to open it up as a contribution to the LangGraph community. It installs from PyPI with pip install fastal-langgraph-toolkit.
Which model providers does it support?
For LLMs, OpenAI including the GPT-5 models, Anthropic, Ollama and AWS Bedrock; for embeddings, OpenAI, Ollama and AWS Bedrock; for speech-to-text, OpenAI Whisper. Switching provider requires no code changes, thanks to configuration injection and the model factory's provider abstraction.
How does its conversation summarization differ from the standard approach?
Automatic summarization driven by token count alone does not hold up on real, complex agents. The method included in the toolkit works on the structure of the conversation instead: it counts Human+AI pairs, excludes ReAct tool calls from the summaries and preserves the context of recent exchanges.
How does it fit into an existing LangGraph graph?
Through the summary_node() method, a ready-to-use node, and through SummarizableState, a type-safe base class that needs no manual field initialization and is natively compatible with LangGraph checkpointing.

Related content

Industries

Services