LangChain Hub

LangChain provides the agent engineering platform and open source frameworks developers need to ship reliable agents fast.

Agent Marketplace

Visit Website

Founded

2022

Location

San Francisco, CA

Employees

150

Funding

Backed by LangChain raise

LangChain Hub (part of LangSmith)

LangChain Hub is a public and private prompt repository built into LangSmith that helps teams discover, test, version, and deploy prompts for LLM applications. It combines a community library with team workspaces, a visual playground, and programmatic pull/push via the LangChain SDK so you can iterate quickly in the UI and ship stable, tagged versions in code.

Visit the Hub: [smith.langchain.com/hub](https://smith.langchain.com/hub)

Launch announcement: [LangChain Prompt Hub](https://blog.langchain.com/langchain-prompt-hub/)

Docs: [Manage prompts](https://docs.langchain.com/langsmith/manage-prompts) and [Programmatic pull/push](https://docs.langchain.com/langsmith/manage-prompts-programmatically)

Pricing: [LangSmith plans](https://www.langchain.com/pricing)

Company: [LangChain on LinkedIn](https://www.linkedin.com/company/langchain)

What LangChain Hub Does

Centralizes prompts for LLM apps with **search, filters, and versioning**.

Lets you browse community prompts, fork to your workspace, edit in a **Playground**, and **commit changes** with full history.

Enables **stable production deployment** by pinning prompt tags; pull by handle and tag from code.

Integrates with LangSmith for **tracing, monitoring, datasets, and evals**, so prompt changes live next to your tests and production telemetry.

How It Works

1. Explore prompts in the Hub (filter by use case, model, language, type).

2. Open any prompt in the **Playground**, add your own model API key, tweak, and commit.

3. Each change creates a **new commit**; tag versions for production.

4. Use the LangChain SDK to **pull** prompts by handle and **push** updates from Python or TypeScript.

Learn more: [Manage prompts](https://docs.langchain.com/langsmith/manage-prompts) and [Programmatic pull/push](https://docs.langchain.com/langsmith/manage-prompts-programmatically).

Access levels:

Read-only browsing without login.

Read/write with a LangSmith login (free Developer plan includes Hub and Playground).

Key Features

Prompt discovery with curated lists (top favorited, viewed, downloaded, recently updated).

Powerful filters for **use case, model family, language, and prompt type**.

**Playground editing** with your own API keys (stored in the browser for model calls).

**Version control** via commits and tags to pin production releases.

**SDK integration** to pull/push prompts in CI/CD.

Deep **LangSmith integration** for evals, datasets, traces, and monitoring.

Notable Prompt Categories and Models

Tags: Chatbots, Agents, RAG, Summarization, Extraction, SQL, Multi-modal. Explore on the [Hub](https://smith.langchain.com/hub).

Model families commonly tagged: OpenAI GPT-3.5/4, Anthropic Claude 2/3, Google Gemini/PaLM, Meta Llama 2/3, and more. Browse examples on the [Hub](https://smith.langchain.com/hub).

Popular Examples

RAG prompt: [rlm/rag-prompt](https://smith.langchain.com/hub/rlm/rag-prompt)

ReAct-style prompt: [hwchase17/react](https://smith.langchain.com/hub/hwchase17/react)

Text-to-SQL: [rlm/text-to-sql](https://smith.langchain.com/hub/rlm/text-to-sql)

OpenAI Functions Agent: [openai-functions-agent](https://smith.langchain.com/hub/hwchase17/openai-functions-agent)

Who It’s For

AI engineers and data scientists building LLM apps who need **shared, versioned prompts**.

Teams that want **fast iteration in a UI** and **stable deployment via tags in code**.

Startups and enterprises already using LangSmith for **tracing, evals, and agent operations**.

Primary Use Cases

Retrieval-augmented generation (RAG) chat and QA (see the [RAG prompt](https://smith.langchain.com/hub/rlm/rag-prompt)).

Agent prompts for tools and function calling (see the [Functions Agent](https://smith.langchain.com/hub/hwchase17/openai-functions-agent)).

Text-to-SQL analytics assistants (see [text-to-sql](https://smith.langchain.com/hub/rlm/text-to-sql)).

Content generation and summarization, e.g., blog generators and YouTube transcript-to-article workflows.

Integrations and Workflow

SDKs: Pull/push prompts in Python and TypeScript via the LangChain SDK. See [Programmatic pull/push](https://docs.langchain.com/langsmith/manage-prompts-programmatically).

Observability: Use LangSmith to tie prompt updates to **traces, datasets, and evals** for measurable improvements. Docs: [Manage prompts](https://docs.langchain.com/langsmith/manage-prompts).

Pricing and Plans

Free: LangSmith Developer plan includes **Prompt Hub** and **Playground** for one seat.

Plus: $39 per seat/month for up to 10 seats with additional deployment features and support.

See details: [LangSmith pricing](https://www.langchain.com/pricing).

Data Handling

Playground uses your own model API keys, which are **stored in the browser** for calls. See the launch post: [Prompt Hub overview](https://blog.langchain.com/langchain-prompt-hub/).

What Users Like

Rapid prototyping and iteration using the UI and SDK; faster experimentation and cleaner assistant architecture. See [G2 reviews](https://www.g2.com/products/langchain/reviews).

Large community library across many models, accelerating baseline performance. Browse the [Hub](https://smith.langchain.com/hub) and see community discussions like this [prompt hub purpose thread](https://www.reddit.com/r/LangChain/comments/1o3zfqo/need_help_understanding_purpose_of_hub/).

Integrated versioning and tags suited for production with LangSmith evals and tracing. Learn more in the docs: [Manage prompts](https://docs.langchain.com/langsmith/manage-prompts).

Considerations and Limitations

Some users cite **UI churn and stability issues** in LangSmith over time. Example discussion: [UI concerns](https://www.reddit.com/r/LangChain/comments/1h84qim/is_langsmith_just_good_piece_of_trash/).

Occasional confusion around **prompt management at scale** or pulling behavior; see troubleshooting threads: [pull issues](https://www.reddit.com/r/LangChain/comments/1aqaejb/not_sure_why_langsmith_hub_doesnt_work/) and [workflow struggles](https://www.reddit.com/r/LangChain/comments/1ctk5l5/struggling_with_prompt_management_tools/).

Broader critiques of LangChain include **fast-moving interfaces and documentation gaps**, impacting onboarding for some teams. Context: [article](https://www.designveloper.com/blog/is-langchain-bad/) and [community thread](https://www.reddit.com/r/LangChain/comments/1gmfyi2/why_are_people_hating_langchain_so_much/).

Public prompts are **community-contributed and unreviewed**; use as starting points and validate before shipping.

Getting Started

Browse and fork prompts on the [Hub](https://smith.langchain.com/hub).

Edit in the **Playground**, commit changes, and tag a release.

Pull the tagged prompt in code using the LangChain SDK. See the docs: [Programmatic pull/push](https://docs.langchain.com/langsmith/manage-prompts-programmatically).

Quick Facts

Product: LangChain Hub (within LangSmith) — [Hub homepage](https://smith.langchain.com/hub)

Core value: Discover, test, version, and manage prompts for LLM apps — [overview](https://blog.langchain.com/langchain-prompt-hub/)

Access: Read-only without login; read/write with LangSmith login

Versioning: Commits with tags for production pinning — [docs](https://docs.langchain.com/langsmith/manage-prompts)

SDK actions: Pull and push prompts from code — [programmatic guide](https://docs.langchain.com/langsmith/manage-prompts-programmatically)

Models: OpenAI, Anthropic, Google, Meta, and more — [browse models via tags](https://smith.langchain.com/hub)

Company: LangChain — [LinkedIn](https://www.linkedin.com/company/langchain)

Related Companies

AI Agent Store

Baseplate

Teams use Baseplate to power their LLM Apps. With Baseplate, you can build scalable backend systems that allow LLMs like GPT-4 to work with your data, without ever being trained on your private information.

ChatHub

Hugging Face Spaces

The AI community building the future.

LobeHub

Metaschool Agent Hub

The home for builders to learn AI & Web3. We create fun learning projects for builders that become the legos for building successful products. We are always looking for ambitious and hungry people to join our team.