Recursiv Docs

How Does Recursiv Compare to LangChain?

LangChain is a library you import to compose chains, agents and tool calls. You still bring your own infrastructure: where it runs, where memory lives, how it deploys and how it is governed. Recursiv is the runtime and the backend. Orchestration, memory, deploy and govern are built, not assembled.

The differentiator is that Recursiv runs the agents. It is not only a set of building blocks you import into an app you operate. The platform schedules work, isolates execution in sandboxes, persists memory, deploys and records what happened, so you can prove what ran.

Quick Comparison

Capability	Recursiv	LangChain
Type	Platform and runtime	Library
Agent composition	Agents and swarms	Chains and agents
Model support	Any model, swap anytime	Any model via integrations
Memory	Brain and memory built in	Bring your own store
Orchestration	Dispatcher and jobs	Bring your own runtime
Code execution	Managed sandboxes	Bring your own
Where it runs	The platform runs it	Your infrastructure
Deploy	Deployments built in	Bring your own host
Data and storage	Postgres and object storage	Bring your own
Auth	Built in	Bring your own
Govern	Scoped identity, approvals, audit, self-eval	Build it yourself
Integrations	Hundreds built in	Many integrations

When LangChain Is the Better Choice

You want a library to drop into an existing app and you already own your infrastructure
You need fine-grained control over chain composition and prompt plumbing in your own code
You are prototyping locally and do not yet need a managed runtime, memory or govern
Your stack is fixed and you want a framework that adapts to it rather than a platform to adopt

Get Started

If you want the orchestration, memory, deploy and govern built rather than assembled, start with the quickstart.