LangChain once held promise as a go-to framework for many developers to build applications powered by LLMs. Even then, it was not perfect and people had a lot of issues. However, a growing number of developers are now moving away from it, citing issues ranging from unnecessary complexity to unstable updates.
While some still find value in LangChain’s features, the overall sentiment suggests that many seek alternatives such as Pydantic or LlamaIndex. One of the most common complaints among developers is LangChain’s instability. Frequent changes to the API structure, coupled with inconsistent documentation, have frustrated users.
In a Reddit discussion, a developer said, “It’s unstable, the interface constantly changes, the documentation is regularly out of date, and the abstractions are overly complicated.” Similar sentiments are echoed throughout the community. Many developers find themselves reading the source code instead of relying on the documentation.
‘LangChain is Overcomplicating Things for No Reason’
A few months back, the engineering team at Octomind, a software company, wrote a detailed blog on why they dropped out of LangChain. The framework’s inflexibility made it difficult to improve lower-level behaviour, and its intentional abstraction of details hindered writing lower-level code.
“When we wanted to move from an architecture with a single sequential agent to something more complex, LangChain was the limiting factor,” read the blog.
LangChain’s complexity has led many to question its design choices. Developers have criticised its layers of abstraction, which make it harder to understand and modify. Experienced developers like Praveer Kochhar, co-founder of Kogo Tech Labs, have questioned the framework and declared that it is not meant for production.
Meanwhile, Angelina Y, the co-founder of OSCR AI, said that as time passes, more people realise that frameworks like LangChain and LlamaIndex are not good for production. “Practically becoming a versatile tool of no use! Of course, I must say that they are very good for making prototypes, especially LlamaIndex,” she added.
Many feel that the framework prioritises “enterprise-level” aesthetics over practical usability.
Last year, AIM also noted that there are a lot of problems with LangChain that continue to remain unresolved. It also uses the same amount of code as the original libraries of OpenAI and others, which makes it feel like bloatware on top of the original APIs, making it inefficient for production use.
For a framework that aims to help developers build reliable AI applications, many find LangChain unsuitable for production. A developer said that their team did a POC project with LangChain, and there were so many changes that they couldn’t update without major code edits. “We are going to get rid of LangChain in our code instead of upgrading it.”
While some developers acknowledge that LangChain is still in rapid development, many feel it lacks the stability required for serious projects. While LangGraph, a related project, is stable, LangChain itself has become bloated.
No Other Alternative
Kieran Klaassen, co-founder of Every Inc, said, “LangChain is where good AI projects go to die.” He added that experienced developers call it “the worst library they’ve ever worked with” due to its bloated abstractions and black-box design.
He advised developers to build their own stack instead. “You’ll spend less time fighting someone else’s broken framework and more time shipping actual features that work.”
Given these challenges, many developers are exploring alternatives that are, admittedly, also not there. Even then, some prefer custom-built solutions over relying on an unstable framework.
For example, PydanticAI offers a more streamlined approach and is ‘Pythony’. This seems similar to what LangChain was known for — the PyTorch for building LLMs. However, just like LangChain, PydanticAI also faces similar issues.
Another emerging alternative is PocketFlow, which aims to provide a more modular and developer-friendly experience. Developers have also opted for LlamaIndex for a long time.
While LangChain has its proponents, the growing dissatisfaction suggests it must address key concerns to regain developer trust. Stability, better documentation, and a focus on practical usability over unnecessary abstractions could help prevent further decline.
However, for many, the damage may already be done. While it may still be useful for rapid prototyping, many are moving to more stable and flexible alternatives. Whether LangChain can turn things around remains to be seen. For now, however, many developers are letting it go.
