LangGraph is hands-down the GOAT of Agentic frameworks! 🐐🔥 A deep dive is brewing, so hit that subscribe button and don’t miss my live stream! 🚀 P.S. Smash that like button if you think my editor deserves to be paid-let’s make it happen! 💸✂
I love the UI and the way langraph works but in practice with tools either PydanticAI or SmolAgents were 100x easier to set-up and get it working properly in like 1/4 of the lines vs langraph. What do you think about those 2 packages I mentioned?
How lol there is no difference. U just prefer that. Langgraph will stay the king and will continue to develop faster and further than the rest. They are the only ones doing updates daily… but if “easier” is what u need 😂
You are absolutely correct. Langgraph requires a lot more boilerplate and code to setup. I have a video on PydanticAI as well. Absolutely love it that tool! I think there is no one framework to rule em all out. I would stick to using a combination of these. For example: PydanticAI for structured response, LangGraph when I needs some complex, cyclic orchestration patterns.
@@YourTechBudCodes have you looked into Instructor ? I would love to use PydanticAI (with LangGraph for more complex apps) but its not really structuring the output (?) just type-hinting and retrying when it fails, unlike Instructor which uses the openAI-structured API that doesn't need retries. Would love some feedback.
LangGraph is hands-down the GOAT of Agentic frameworks! 🐐🔥 A deep dive is brewing, so hit that subscribe button and don’t miss my live stream! 🚀
P.S. Smash that like button if you think my editor deserves to be paid-let’s make it happen! 💸✂
SUBSCRIBED !! We need more video on developing AI agents, keep it up !!
Thats the plan! Glad you are finding this useful!
@@YourTechBudCodes Bro please keep this going , it's really helpful thanks a lot man , i turned on the bail notification as well🥰
I love the UI and the way langraph works but in practice with tools either PydanticAI or SmolAgents were 100x easier to set-up and get it working properly in like 1/4 of the lines vs langraph. What do you think about those 2 packages I mentioned?
How lol there is no difference. U just prefer that. Langgraph will stay the king and will continue to develop faster and further than the rest. They are the only ones doing updates daily… but if “easier” is what u need 😂
You are absolutely correct. Langgraph requires a lot more boilerplate and code to setup. I have a video on PydanticAI as well. Absolutely love it that tool! I think there is no one framework to rule em all out. I would stick to using a combination of these. For example: PydanticAI for structured response, LangGraph when I needs some complex, cyclic orchestration patterns.
@@YourTechBudCodes Hey I am new with langgraph whats the case with structured output in langgraph. How do you compare it with pydanticai.
@@YourTechBudCodes have you looked into Instructor ? I would love to use PydanticAI (with LangGraph for more complex apps) but its not really structuring the output (?) just type-hinting and retrying when it fails, unlike Instructor which uses the openAI-structured API that doesn't need retries. Would love some feedback.
If had that capability, i would have give 1 million likes 👍
You are really sweet. Thank you!