Latest publications
LangChain Introduced Interpreter Skills to Expand Agent Capabilities
LangChain has added a new way to expand AI agents through Interpreter Skills—TypeScript modules that agents can import and execute to solve complex tasks.
LangChain Releases Mission Control for Managing LangSmith on Kubernetes
LangChain introduced Mission Control, a tool for operational management of self-hosted LangSmith. It helps with cluster configuration, health checks, releases, and diagnostics on Kubernetes.
Interrupt 2026: LangChain Unveiled Tools for Debugging AI Agents in Production
LangChain held a two-day Interrupt 2026 conference where it presented new tools for diagnosing and debugging AI agents in production environments.
LangSmith Released Sandboxes for Secure Execution of Coding Agents
LangSmith announced the release of Sandboxes — an isolated environment based on kernel-isolated micro-VMs for secure execution of code generated by AI agents.
Lyft reduced AI-agent development from months to weeks with LangGraph and LangSmith
Lyft created a platform that allows teams to quickly develop and deploy AI agents for customer support using LangGraph and LangSmith.
How LangChain Protected Agents in LangSmith from Credential Leaks
LangChain added Auth Proxy to LangSmith Sandboxes — an access control system that hides API keys from agents and restricts their outbound connections.
LangChain Moves from Token Streaming to Agent Streams
LangChain has introduced streaming primitives for typed events and sub-agent visibility, enabling developers to build production-ready AI agents with reliable frontend experience.
Agent Harness in LangChain: the architecture of autonomous AI assistants
Agent Harness turns AI models into autonomous workers through three key components: file systems for data access, isolated sandboxes for security, and memory for context.
LangChain Launches Engine — Automatic Diagnostics for Agent Errors
LangSmith Engine automatically monitors production agents, groups errors into named issues, and suggests targeted fixes instead of manual log analysis.
LangChain Introduced SmithDB — A Distributed Database Built for AI Agents
LangChain introduced SmithDB, a distributed database for tracking AI agent operations. It is 12 times faster than alternatives and fully portable, deployable on any infrastructure.
LangSmith Introduces LLM Gateway: Cost Control and PII Protection for AI Agents
LangSmith has added a built-in LLM Gateway to manage costs and security of LLM requests throughout the AI agent lifecycle—with spend limits, PII removal, and full call tracing.
LangChain optimized Deep Agents for different models: +10–20% performance gain
Deep Agents now have model-specific profiles that tailor behavior to OpenAI, Anthropic, and Google models. On the tau2-bench benchmark, this delivered a 10–20 point performance improvement.
Deep Agents 0.6: update to LangChain's agent framework
LangChain has released Deep Agents 0.6 with a code interpreter, configuration profiles, and performance optimizations. Key additions include streaming v3, delta channels, and ContextHub for managing agent memory.
LangChain Introduces DeltaChannel to Save Memory for Long-Lived Agents
The new DeltaChannel primitive in LangGraph 1.2 solves the problem of exponential memory growth during long agent sessions by storing only changes between steps instead of full state.
LangChain Unveiled Automatic Debugging and One-Line Deployment at Interrupt 2026
LangChain released a toolkit for automatic debugging and deployment of AI agents, presented at the Interrupt 2026 conference.
LangChain Launches LangChain Labs to Develop Self-Learning AI Agents
LangChain has created a new research division, LangChain Labs, focused on continuous learning of AI agents and development of systems capable of self-improvement.
LangChain Adds Interpreters to Deep Agents — Code Management Between Tool Calls
LangChain expanded Deep Agents with built-in interpreters, where agents write code to coordinate tools and control model context.