What are the best alternatives to LangChain?

CrewAI is the strongest direct alternative if your problem is multi-agent: it is a "fast and flexible multi-agent automation framework" with two explicit primitives — role-based Crews for autonomy, event-driven Flows for control — and it states it is standalone rather than a layer on something else. AutoGen was the other obvious answer and is no longer one for new projects: its README puts it in maintenance mode and points new users at Microsoft Agent Framework. If what you actually want is orchestration rather than a different framework, LangChain's own answer is LangGraph, which its README describes as the framework "for building controllable agent workflows".

Is AutoGen deprecated?

Its own README describes it as being in maintenance mode, which is close to but not the same as deprecated: "AutoGen is now in maintenance mode. It will not receive new features or enhancements and is community managed going forward. New users should start with Microsoft Agent Framework. Existing users are encouraged to migrate using the AutoGen → Microsoft Agent Framework migration guide." So it still installs, it is still MIT (alongside a CC-BY-4.0 licence for non-code content), and the code you already depend on has not been withdrawn — but nothing new is coming, and Microsoft has said where to go instead. That is a far better position to be in than an unmaintained project that says nothing.

Which agent framework is most actively maintained?

That is a question about this week rather than about the documentation, and none of the three publishes an answer. The table above records what each project says about itself; StillShipping's board records what each repository actually did, rescored nightly from the GitHub API on commit recency, release cadence, issue response time and maintainer count. AutoGen is the only one of the three whose README says anything about its own status at all. Treating the other two's silence as evidence of health is the specific mistake this site exists to make unnecessary.

Do any of these lock you into a paid product?

Not at the licence, and it is worth knowing where the money is anyway. All three are MIT-licensed. LangChain sells LangSmith — "agent evals, observability, and debugging for LLM apps" — and a deployment platform beside it. CrewAI sells the AMP Suite, "a commercial control plane around CrewAI" adding managed deployment, observability, governance, security and enterprise support. AutoGen publishes no paid tier at all, which in its case is less reassuring than it sounds: what funds a free framework is usually the paid product beside it, and AutoGen's successor is the thing Microsoft is now investing in.

How is this list different from a roundup article?

A roundup is a snapshot of the day someone wrote it, and the thing most likely to have changed since is not the API surface — it is whether anyone is still merging. Every claim in the table above is a quotation from the project's own repository with the date it was read, and every figure on the board is measured from the GitHub API on the night you are reading. Where a project publishes nothing on an axis, the cell says "Not published" and names the page that was read rather than inferring an answer — which is how a table stays a reference rather than becoming an opinion.

Board / 3 compared, every cell read off its own repository on September 18, 2026, rescored September 13, 2026

LangChain alternatives in 2026

Three multi-agent frameworks on the same six criteria, every claim read off that project's own repository. Beside the freshness score this site measures every night.

The verdict

what to pick, and what the answer rests on

Start with the one fact none of the roundups carries: AutoGen has stopped. Its own README says so. "AutoGen is now in maintenance mode. It will not receive new features or enhancements and is community managed going forward". And Microsoft names the successor, Microsoft Agent Framework, with a migration guide. If you are choosing today, that removes a column. Between the other two, choose CrewAI when the work is a team of specialists and you want one framework covering both autonomy and control, and choose LangChain when integration surface and ecosystem matter more than opinion. Both are MIT, both are company-owned, both sell operational tooling beside the framework. LangSmith and the CrewAI AMP Suite. And neither publishes a word about its own maintenance. That last point is the general case rather than an accusation: AutoGen is the exception here, not the rule, and a reader who takes silence for health is reading nothing at all. StillShipping reads those scores from the GitHub API every night for exactly that reason, and publishes them per tool at /is/langchain-maintained, /is/crewai-maintained and /is/autogen-maintained.

Every column, on every axis

what each one publishes about itself, read off its own repository on September 18, 2026

LangChain alternatives in 2026, on what each project publishes about itself
Axis	LangChain	CrewAI	AutoGen
What it is, in its own words	A framework for building agents and LLM applications"The agent engineering platform. LangChain is a framework for building agents and LLM-powered applications. It helps you chain together interoperable components and third-party integrations."read September 18, 2026	A multi-agent automation framework"Fast and flexible multi-agent automation framework. CrewAI is an open-source Python framework with high-level abstractions and low-level APIs for building production-ready multi-agent workflows."read September 18, 2026	A framework for multi-agent AI applications"AutoGen is a framework for creating multi-agent AI applications that can act autonomously or work alongside humans."read September 18, 2026
What its README says about its own maintenancewhether the project's front page tells a reader that it is still being developed, has stopped, or has a successor. As distinct from what its commit graph implies	Not publishedThe README carries no statement about the project's maintenance status, no deprecation notice and no successor. What it does carry is routing advice within its own ecosystem. "just getting started? check out deep agents" and, for orchestration, LangGraph. Read on the repository's own front page on 14 August 2026.read September 18, 2026	Not publishedNo maintenance statement, deprecation notice or successor anywhere on the README. The closest thing to a health signal it publishes is adoption. "with over 100,000 developers certified through our community courses". Which is a different claim. Read on the repository's own front page on 14 August 2026.read September 18, 2026	Maintenance mode, with a named successor"⚠️ Maintenance mode. AutoGen is now in maintenance mode. It will not receive new features or enhancements and is community managed going forward. New users should start with Microsoft Agent Framework. Existing users are encouraged to migrate using the AutoGen → Microsoft Agent Framework migration guide." The only one of the five that publishes its own status at all.read September 18, 2026
How you install it	uv add langchainThe quickstart is three lines: install, `init_chat_model`, invoke. A JS/TS equivalent, langchain.js, is published separately.read September 18, 2026	uv tool install crewai"Q: How do I install CrewAI? A: Install the CrewAI CLI with UV: uv tool install crewai." The README then says to "create a project with `crewai create crew <project_name>`, run `crewai install`, and start it with `crewai run`." The `crewai[tools]` extra the README used to name is gone from it entirely.read September 18, 2026	pip install autogen-agentchat"AutoGen requires Python 3.10 or later." Installed as `autogen-agentchat` plus an extension package for the model client; `autogenstudio` is published separately as a no-code GUI.read September 18, 2026
Licence	MITShown in the repository's own file navigation alongside its code of conduct, contributing guide and security policy.read September 18, 2026	MIT"CrewAI is released under the MIT License." Stated in the README as well as in the repository's file navigation.read September 18, 2026	MIT, plus CC-BY-4.0The repository's file navigation lists two: a CC-BY-4.0 licence and an MIT licence, with a separate LICENSE-CODE file. The only dual-licensed project of the five.read September 18, 2026
Who publishes it	langchain-aiA company-owned repository, which is the arrangement four of these five share and the one that most affects what happens to the project if the company's priorities move.read September 18, 2026	crewAIIncCompany-owned, with a commercial suite published from the same README. So the framework and the product it feeds are maintained by the same organisation.read September 18, 2026	microsoftPublished from Microsoft's own organisation, which is also what makes its maintenance notice load-bearing: the successor it names is another Microsoft product.read September 18, 2026
What it sells alongside	LangSmith, and a deployment platform"LangSmith - agent evals, observability, and debugging for LLM apps" and "LangSmith Deployment - deploy and scale agents with a purpose-built platform". The framework is MIT; the paid surface sits beside it rather than inside it.read September 18, 2026	CrewAI AMP Suite"For organizations that need a commercial control plane around CrewAI, CrewAI AMP Suite adds managed deployment, observability, governance, security, and enterprise support." One part, the Crew Control Plane, is offered free to try.read September 18, 2026	Not publishedThe README publishes no paid tier, hosted plan or commercial suite for AutoGen itself. What it names instead is a successor framework rather than a product sold alongside this one. Read on the repository's own front page on 14 August 2026.read September 18, 2026
Why you would pick itEach project’s own case, in its own words.	You want the largest integration surface in this market and you expect the underlying models to keep changing under you. LangChain's own framing is exactly that. It helps you "chain together interoperable components and third-party integrations to simplify AI application development - all while future-proofing decisions as the underlying technology evolves". And it is the only one of the five that publishes a whole ecosystem around the core: Deep Agents for common patterns, LangGraph for low-level orchestration, LangSmith for evals and observability, and a JS/TS library.	The shape of your problem is a team of specialists rather than a chain of calls, and you want one framework that does both autonomy and control without a second dependency. CrewAI splits those explicitly. "CrewAI Crews: optimize for autonomy and collaborative intelligence with role-based AI agents" and "CrewAI Flows: build event-driven automations that combine precise workflow control, single LLM calls, and native support for Crews". It is also the only one of the five that states it is standalone: "CrewAI is a standalone Python framework with its own primitives for agents, tasks, crews."	You already have AutoGen in production and you are deciding whether to move. Then read its own README first, because it answers the question for you: it is in maintenance mode, it "will not receive new features or enhancements", and Microsoft names the successor. "new users should start with Microsoft Agent Framework", with a published migration guide. A framework that tells you plainly it has stopped is more useful to a maintainer than one that has quietly stopped and says nothing, and that candour is a real reason to trust the code you already depend on while you migrate off it.

What the board says tonight

the axis the table above cannot carry, read from the GitHub API on the same nightly pass

The tools in this comparison, as this board scored them

langchainlangchain-ai/langchain	MaintainedOrchestrators	100/100	0d agoSeptember 12, 2026	100+	22	too few	146k
CrewAIcrewAIInc/crewAI	MaintainedOrchestrators	100/100	1d agoSeptember 11, 2026	100+	31	too few	58k
AutoGenmicrosoft/autogen	SlowingOrchestrators	35/100	5mo agoApril 15, 2026	0	0	too few	61k

Nothing on this row is quoted from anybody. It is the GitHub API, read last night, and it moves without this page being edited.

Questions

5 asked about this comparison

What are the best alternatives to LangChain?: CrewAI is the strongest direct alternative if your problem is multi-agent: it is a "fast and flexible multi-agent automation framework" with two explicit primitives. Role-based Crews for autonomy, event-driven Flows for control. And it states it is standalone rather than a layer on something else. AutoGen was the other obvious answer and is no longer one for new projects: its README puts it in maintenance mode and points new users at Microsoft Agent Framework. If what you actually want is orchestration rather than a different framework, LangChain's own answer is LangGraph, which its README describes as the framework "for building controllable agent workflows".
Is AutoGen deprecated?: Its own README describes it as being in maintenance mode, which is close to but not the same as deprecated: "AutoGen is now in maintenance mode. It will not receive new features or enhancements and is community managed going forward. New users should start with Microsoft Agent Framework. Existing users are encouraged to migrate using the AutoGen → Microsoft Agent Framework migration guide." So it still installs, it is still MIT (alongside a CC-BY-4.0 licence for non-code content), and the code you already depend on has not been withdrawn. But nothing new is coming, and Microsoft has said where to go instead. That is a far better position to be in than an unmaintained project that says nothing.
Which agent framework is most actively maintained?: That is a question about this week rather than about the documentation, and none of the three publishes an answer. The table above records what each project says about itself; StillShipping's board records what each repository actually did, rescored nightly from the GitHub API on commit recency, release cadence, issue response time and maintainer count. AutoGen is the only one of the three whose README says anything about its own status at all. Treating the other two's silence as evidence of health is the specific mistake this site exists to make unnecessary.
Do any of these lock you into a paid product?: Not at the licence, and it is worth knowing where the money is anyway. All three are MIT-licensed. LangChain sells LangSmith. "agent evals, observability, and debugging for LLM apps". And a deployment platform beside it. CrewAI sells the AMP Suite, "a commercial control plane around CrewAI" adding managed deployment, observability, governance, security and enterprise support. AutoGen publishes no paid tier at all, which in its case is less reassuring than it sounds: what funds a free framework is usually the paid product beside it, and AutoGen's successor is the thing Microsoft is now investing in.
How is this list different from a roundup article?: A roundup is a snapshot of the day someone wrote it, and the thing most likely to have changed since is not the API surface. It is whether anyone is still merging. Every claim in the table above is a quotation from the project's own repository with the date it was read, and every figure on the board is measured from the GitHub API on the night you are reading. Where a project publishes nothing on an axis, the cell says "Not published" and names the page that was read rather than inferring an answer. Which is how a table stays a reference rather than becoming an opinion.

Where every cell was read

derived from the cells themselves, never listed by hand

Source	Read	Why
LangChain	September 18, 2026	What it is, in its own words, What its README says about its own maintenance, How you install it, Licence, Who publishes it, What it sells alongside. github.com/langchain-ai/langchain
CrewAI	September 18, 2026	What it is, in its own words, What its README says about its own maintenance, How you install it, Licence, Who publishes it, What it sells alongside. github.com/crewAIInc/crewAI
AutoGen	September 18, 2026	What it is, in its own words, What its README says about its own maintenance, How you install it, Licence, Who publishes it, What it sells alongside. github.com/microsoft/autogen

A hand-listed source block drifts from the table the first time a cell’s URL changes, and then the page is citing a page it did not read.