Deep Agents is actively used in production by OpenSWE and LangSmith Fleet.
Overview
| Aspect | LangChain Deep Agents | Claude Agent SDK | Codex SDK |
|---|---|---|---|
| Use cases | Custom general-purpose agents (including coding) | Custom general-purpose agents (including coding) | Prebuilt coding agent that can execute coding tasks |
| Model support | Flexible and model-agnostic (Anthropic, OpenAI, and 100s others) | Tightly integrated with Claude models (Anthropic, Azure, Vertex AI, AWS Bedrock) | Tightly integrated with OpenAI models (GPT-5.3-Codex and variants) |
| Architecture | Python SDK, TypeScript SDK, and CLI | Python SDK, TypeScript SDK | TypeScript SDK, CLI, desktop app, IDE extension, cloud interface |
| Execution environment | Local, remote sandboxes, virtual filesystem | Local, cloud containers | Local, cloud |
| Deployment | deepagents deploy | Self-hosted | N/A |
| Frontend | Integration with React | Server-side only | Server-side only |
| Observability | LangSmith tracing & evaluations | N/A | OpenAI traces & command-line evaluations |
| Security configurability | Composable, per-tool human-in-the-loop | Permission system with modes, rules and hooks | Built-in tiers using approval modes and OS-level sandboxes |
| License | MIT | MIT (underlying Claude Code is proprietary) | Apache-2.0 |
Summary
The central question is deployment. If you are building an agent that will serve multiple users in production, Deep Agents gives you auth, multi-tenant isolation, open protocol endpoints, and observability out of the box. Claude Agent SDK is built on top of Claude Code, a local development tool, and requires you to build and maintain those layers yourself.- Choose Deep Agents to go to production with one command: a model-agnostic, multi-tenant deployment with built-in auth, observability, and open protocol endpoints, without assembling your own hosting stack, while keeping control of your memory.
- Choose Claude Agent SDK to standardize on Claude models and build your own production infrastructure. It is well-suited for local development and custom server architectures.
- Choose Codex SDK to standardize on OpenAI models with a prebuilt coding workflow and cloud execution environment.
Feature comparison
The feature comparison is grouped by decision area.Deployment and production
| Feature | Deep Agents | Claude Agent SDK | Codex SDK |
|---|---|---|---|
| Production hosting | ✅ LangSmith Deployment, deepagents deploy, self-hosted, or standalone | Build your own HTTP/WebSocket layer; run SDK in containers | ❌ |
| Multi-tenant deployment | ✅ One deployment with custom auth and scoped user data, plus RBAC | ⚠️ Possible, but you must build auth, tenancy, and per-user isolation yourself | ❌ |
| Local execution environment | ✅ | ✅ | ✅ |
| Cloud execution environment | ❌ | ❌ | ✅ |
Need user-level isolation in one shared deployment? Deep Agents support multi-tenant deployments with scoped threads, runs, and conversation history. For a walkthrough, watch the custom auth video.
Agent capabilities
| Feature | Deep Agents | Claude Agent SDK | Codex SDK |
|---|---|---|---|
| File Read/Write/Edit | ✅ read_file, write_file, edit_file | ✅ Read, write, edit | ✅ Read, write, edit |
| Shell Execution | ✅ execute | ✅ bash | ✅ exec |
| Glob/Grep | ✅ glob, grep | ✅ glob, grep | ✅ Built-in |
| Web Search | ✅ Support for third-party and provider-native | ✅ WebSearch, WebFetch | ✅ web_search |
| Planning/Todos | ✅ write_todos | ✅ Todo lists | ✅ Plan before changes |
| Subagents | ✅ Subagents | ✅ Subagents | ✅ Multi-agent workflows (experimental) via MCP + Agents SDK |
| MCP client (connect external tools) | ✅ | ✅ MCP | ✅ MCP client |
| Human-in-the-Loop | ✅ Approve/edit/reject | ✅ Permission modes | ✅ Approval modes (Full Access, Read-only, Auto) |
| Skills | ✅ Skills | ✅ Skills | ✅ Skills |
| Agent instructions file (loaded at startup) | ✅ Memory | ✅ CLAUDE.md files | ✅ AGENTS.md |
| Long-term memory (persist context across sessions) | ✅ Memory Store | ❌ | ❌ |
| Streaming | ✅ Streaming | ✅ Streaming | ✅ Streaming |
Sandboxing and architecture
| Feature | Deep Agents | Claude Agent SDK | Codex SDK |
|---|---|---|---|
| Agent runs in sandbox | ✅ Can run in sandbox | ✅ Can run in sandbox | ✅ Can run in sandbox |
| Agent runs operations in sandboxes | ✅ Uses sandboxes as tools | ❌ | ❌ |
| OS-level sandbox modes | ❌ | ❌ | ✅ read-only, workspace-write, and danger-full-access |
| Middleware / lifecycle hooks | ✅ | ✅ Hooks | ❌ |
| Virtual filesystem (pluggable storage backends) | ✅ Virtual filesystem with pluggable backends | ❌ | ❌ |
Protocols and state
| Feature | Deep Agents | Claude Agent SDK | Codex SDK |
|---|---|---|---|
| ACP server (expose agent to IDEs) | ✅ ACP server | ✅ (third-party) | ✅ (third-party) |
| MCP server mode (expose agent as a tool) | ✅ MCP endpoint through Agent Server | ❌ | ✅ codex mcp-server |
| A2A endpoint (agent-to-agent protocol) | ✅ A2A endpoint in Agent Server | ✅ (third-party) | ❌ |
| Session resume (reload conversation history) | ✅ Reconnect and load thread history | ✅ Session management | ✅ Resume threads; SDK resumeThread() |
| Checkpoint to disk (resume across restarts) | ✅ Backends and checkpoints | ✅ File checkpointing | ✅ Git checkpoints |
| State branching (fork and replay sessions) | ✅ | ✅ | ✅ |
Observability
| Feature | Deep Agents | Claude Agent SDK | Codex SDK |
|---|---|---|---|
| Native Tracing | ✅ LangSmith | ❌ | ✅ OpenAI Traces |
| Evaluations | ✅ LangSmith | ❌ | Evals patterns (command-line with codex exec --json) |
Notice a mistake?We drafted this comparison on March 4th, 2026.
If you notice mistakes or changes in products, please file an issue.
Connect these docs to Claude, VSCode, and more via MCP for real-time answers.