We stopped using both about 6 months ago. Here's what replaced them:
For 80% of agent use cases, frameworks are the wrong abstraction. You don't need an orchestration layer — you need a well-structured prompt, persistent memory files, and a cron scheduler. That's it. LangChain and CrewAI abstract over this, but the abstraction adds complexity without adding capability for most projects.
The cases where they still make sense: true multi-agent pipelines where you need agent-to-agent communication, observability, and eval tooling. If you're building a production system where 10+ agents are coordinating and you need to trace what each one did — LangChain's observability hooks are genuinely useful.
For everything else (a solo agent running a business, a writing assistant, a support bot), the overhead of learning and maintaining a framework isn't worth it. Raw API calls + good prompts + file-based memory gives you 90% of the capability at 20% of the complexity.
The real problem with both frameworks: they were designed when LLMs were less capable, so they over-engineer around the LLM's limitations (poor instruction following, limited context windows). Modern models don't need as much scaffolding. The frameworks haven't fully caught up to that.
I feel the hype is cooling down. LangChain was great for getting something running in 5 minutes, but the 'abstraction soup' makes debugging a nightmare in production. I'm seeing more people just using the OpenAI/Anthropic SDKs directly or very thin wrappers. It’s better to own your prompts than to hide them behind five layers of library code.
Well, what are you building? It's hard to know if its the right tool without understanding what problem you're trying to solve.
If you need a multi-model setup, have complex agentic workflows, have observability requirements, need to run evals, etc, then they'll make more sense.
The company I work for uses LangChain heavily, but that's because we have fairly complex requirements compared to products which are just incorporating AI as an additional feature for example.
It's probably similar to how Next.js is overused and overhyped – it can be great if you have complex requirements, but if you just need a simple website with a little interactivity it's total overkill.
LangChain seems to get a lot of hate here and I'm not entirely sure why. There's a lot of frameworks which seem to make a relatively simple problem needless complex, I don't feel that way about LangChain personally.
We stopped using both about 6 months ago. Here's what replaced them:
For 80% of agent use cases, frameworks are the wrong abstraction. You don't need an orchestration layer — you need a well-structured prompt, persistent memory files, and a cron scheduler. That's it. LangChain and CrewAI abstract over this, but the abstraction adds complexity without adding capability for most projects.
The cases where they still make sense: true multi-agent pipelines where you need agent-to-agent communication, observability, and eval tooling. If you're building a production system where 10+ agents are coordinating and you need to trace what each one did — LangChain's observability hooks are genuinely useful.
For everything else (a solo agent running a business, a writing assistant, a support bot), the overhead of learning and maintaining a framework isn't worth it. Raw API calls + good prompts + file-based memory gives you 90% of the capability at 20% of the complexity.
The real problem with both frameworks: they were designed when LLMs were less capable, so they over-engineer around the LLM's limitations (poor instruction following, limited context windows). Modern models don't need as much scaffolding. The frameworks haven't fully caught up to that.
I feel the hype is cooling down. LangChain was great for getting something running in 5 minutes, but the 'abstraction soup' makes debugging a nightmare in production. I'm seeing more people just using the OpenAI/Anthropic SDKs directly or very thin wrappers. It’s better to own your prompts than to hide them behind five layers of library code.
Lol, I felt the same and this sort of also led us to build what i ended up founding now
Sometimes, simple PoC that does not require any "Agentic" features
Well, what are you building? It's hard to know if its the right tool without understanding what problem you're trying to solve.
If you need a multi-model setup, have complex agentic workflows, have observability requirements, need to run evals, etc, then they'll make more sense.
The company I work for uses LangChain heavily, but that's because we have fairly complex requirements compared to products which are just incorporating AI as an additional feature for example.
It's probably similar to how Next.js is overused and overhyped – it can be great if you have complex requirements, but if you just need a simple website with a little interactivity it's total overkill.
LangChain seems to get a lot of hate here and I'm not entirely sure why. There's a lot of frameworks which seem to make a relatively simple problem needless complex, I don't feel that way about LangChain personally.
No. They suck.