LangChain Deep Agents
Agent harness from LangChain
Deep Agents is an agent harness from LangChain for long, multi-step tasks, built on LangChain and the LangGraph runtime. It includes filesystem backends for context management, optional planning with todo lists, skills, long-term memory and human approval. For multi-agent work, the main agent delegates via the task tool to subagents with their own fresh context, each returning a final report; custom subagents are defined in the subagents parameter, and async subagents handle long-running or parallel work. The library is available for Python and JavaScript under the MIT license; tracing and deployment optionally run on the paid LangSmith platform.
Good fit if
Focus: Developers building research, coding or analysis agents for long tasks in the LangChain ecosystem
- your project is about Deep research with subagents, Coding agents or Analysis of large document sets
- someone on your team writes code
Probably not if
- nobody on your team writes code (1/5 for use without developers)
Key figures
- Usable without developers
- 1/5
- Scalability
- 4/5
- Integrations
- 4/5
- Versioning
- 3/5
- Observability
- 4/5
Editorial rating on a scale from 1 to 5. Methodology
Fact sheet
- Category
- Developer
- Type
- Agent harness from LangChain
- Licence
- Open Source (MIT)
- Pricing
- Open Source / CloudLibrary without license fee (MIT). Optional LangSmith for tracing and deployment: Developer $0 per seat/month (1 seat, 5,000 traces/month), Plus $39 per seat/month (10,000 traces, one small serverless deployment included), usage-based beyond that; Enterprise on request.
- Deployment
- Cloud, Self-Hosted, Any
- Models
- OpenAI, Anthropic, Google Gemini, OpenRouter, Fireworks, Baseten, Ollama
- RAG
- Yes
- Multi-Agent
- Yes
- SSO, roles (RBAC)
- SSO no, RBAC no
- Latest version
- 0.7.23 (Python, 07.10.2026); JavaScript 1.14.2 (05.10.2026)
- GitHub stars
- 30,105
Key features
- Subagents via the task tool
- Filesystem backends and sandboxes
- Long-term memory via AGENTS.md
- Skills following the Agent Skills standard
- Human in the loop with interrupt_on
Typical use cases
- Deep research with subagents
- Coding agents
- Analysis of large document sets
- Long-running assistants with memory
Market notes
Changes to LangChain Deep Agents by email
We let you know when the status, pricing or license of LangChain Deep Agents changes, together with the weekly market notes.
Implementation
Get LangChain Deep Agents implemented
Free for you. Partners only pay for requests they accept.
Alternatives
Similar systems
Similar in category, features or use cases.
- DeveloperLangGraphLangGraph is a low-level framework from LangChain for modelling stateful agents as a graph of nodes and edges. Unlike pure DAG tools, LangGraph allows cycles, i.e. loops in which an agent repeats steps until a condition is met. Persistence, checkpoints and human approvals are built in; LangSmith Deployment is available for running it in production. For agents that delegate work to subagents, LangChain Deep Agents is built on top of LangGraph.View profile
- DeveloperLangChainLangChain is one of the most widely used open-source frameworks for LLM applications, with integrations for almost every model provider, vector database and tool. Since version 1.0 its agent functionality is built on LangGraph. The many abstractions make getting started demanding, even for experienced developers.View profile
- DeveloperAgent SquadAgent Squad is an open source framework for Python, TypeScript and Swift that distributes requests across multiple specialized agents. A classifier selects the right agent for each message based on agent descriptions and conversation history; the SupervisorAgent additionally coordinates a team of agents in parallel using an agents-as-tools approach. The project was formerly called Multi-Agent Orchestrator, was hosted under awslabs and is now maintained by 2FastLabs. It is licensed under Apache-2.0 and runs locally, in containers or on AWS Lambda.View profile
- DeveloperBeeAI FrameworkThe BeeAI Framework is an open source framework for agents and multi-agent systems in Python and TypeScript, originally developed by IBM and now under open governance at the Linux Foundation (LF AI & Data). The RequirementAgent follows rules the developer defines, such as which tool must be called at which step. Multiple agents collaborate via the HandoffTool, which lets a main agent pass questions to specialists, and via workflows; agents can be served over A2A and MCP. It supports more than ten model providers, OpenTelemetry, memory, RAG and caching. The license is Apache-2.0.View profile