AI BriefingLangChain Blog news
The latest AI and developer news from LangChain Blog, with short summaries.
Visit website →LangSmith adds AI Search to trace agent behavior with natural language
AI Search is available in public beta on LangSmith Cloud at no additional cost, enabling semantic queries across agent traces.
LangChain integrates ScaleDown constrained decoding models for efficient long-running agents
ScaleDown models offer up to 467x cost savings and 200x speed improvements over frontier models for structured tasks like summarization and extraction.
Deep Agents Uses Five Middleware Techniques to Prevent Context Rot in Long-Running LLM Tasks
The open-source harness moves context out of the active window using offloading, summarization, skills, subagents, and long-term memory.
Managed Deep Agents v0.9 introduces Slack reactions API for dynamic loading states
Managed Deep Agents v0.9 introduces a new reactions attribute for Slack channels, allowing developers to dynamically assign emoji responses using simple heuristics or decision models like Jev.
Snyk Moves Internal Support Agent to Customer-Facing Product with 85% Ticket Deflection
The LangGraph-powered agent has handled over 60,000 queries and resolves more than 85% of sessions without creating a support ticket.
LangChain Releases Restock Sample Agent for AI Payments via Stripe Link and MPP
The sample agent runs on Managed Deep Agents and uses MPP-enabled APIs to execute real purchases with user-approved budgets.
LangChain Revamps Skills in Deep Agents with Tool Binding and Runtime Pinning
The update allows tools to load only when a skill is read and enables mid-thread skill reloading without starting a new session.
PickLangChain Releases Managed Deep Agents v0.9 with Schedules, Per-Run Configuration, and Slack Reactions
The new version enables agents to create their own cron schedules, dynamically reconfigure models and tools per run, and acknowledge Slack messages with reactions.
LangChain Releases Model Routing Middleware to Cut Coding Agent Costs by 64%
The new middleware reduced median cost per coding task by 64% in Open SWE experiments without impacting quality.
LangChain Demonstrates Building Production Agents with TypeSafe’s Jev and LangGraph
LangChain details how to use TypeSafe’s Jev decision model with LangGraph to create faster, cheaper agent workflows by reserving LLMs for complex reasoning.
PickLangSmith Launches Engine v2, Managed Deep Agents v0.8, and Fine-Tuning Tools
LangSmith Engine v2 introduces red teaming and automated fix validation, while Managed Deep Agents v0.8 adds user-level memory and built-in web search.
LangSmith Engine v2 Adds Red Teaming and Automated Fix Validation
Engine v2 proactively detects issues like hallucinations and validates fixes before deployment, with new features available in Private Beta for LangSmith Deployment users.
Managed Deep Agents 0.8 Launches with User-Scoped Memory, HTTP Channels, and Parallel Web Search
LangChain has released Managed Deep Agents v0.8, introducing user-level memory isolation, HTTP channels for integration, and built-in web search powered by Parallel.
PickLangSmith Launches Trajectories for Readable Agent Session Debugging
Trajectories project nested traces into a linear, readable view of agent actions, enabling faster debugging and SME review.
LangChain Launches LangSmith Fine-Tuning CLI for Agent Trajectory Post-Training
The new `smithtune` CLI enables end-to-end supervised fine-tuning using agent trajectories, partnering with Fireworks and Baseten for training infrastructure.
LangSmith Launches Custom Apps for Building Tailored Interfaces on Agent Data
LangSmith has launched Custom Apps, allowing Plus and Enterprise users to build and publish custom interfaces on agent data, with Plus plans including one app and chat-based creation.
LangSmith Used as Clinical Evaluation Infrastructure to Ensure Reliability in Medical AI
By leveraging LangSmith to transform clinical expert judgments into reusable evaluation assets, teams have simultaneously secured release speed and safety for medical AI.
LangSmith Adds Jev Evaluation Model… Agent Evaluation Up to 450x Cheaper and 200x Faster Than LLMs
LangSmith has integrated TypeSafe AI's Jev model to support agent evaluations that are up to 450x cheaper and 200x faster than LLMs.
PickLangChain Reveals Experimental Results for TypeSafe AI's 'Jev': Up to 200x Faster and More Consistent than LLM Judges
LangChain's validation of TypeSafe AI's Jev confirmed it is up to 200x faster, 400x cheaper, and more consistent than LLM Judges for classification tasks, though the experiment is in its early stages and requires further verification.
PickTypeSafe AI Unveils 'Jev', a Structured Decision Model 200x Faster Than LLMs
TypeSafe AI has released 'Jev', a structured decision model that is 200x faster and 400x cheaper than LLMs.