# LLM Response Error: ValueError: Invalid response from LLM call - None or empty
**URL:** https://community.crewai.com/t/llm-response-error-valueerror-invalid-response-from-llm-call-none-or-empty/4425
**Category:** LLMs
**Tags:** llama-31-8b
**Created:** [March 6, 2025, 6:07am UTC](https://community.crewai.com/t/llm-response-error-valueerror-invalid-response-from-llm-call-none-or-empty/4425 "2025-03-06T06:07:10Z")
**Posts on this page:** 6
**Page:** 1
### Author:  [@sranjan710](https://community.crewai.com/u/sranjan710)
#### Post date: [March 6, 2025, 6:07am UTC](https://community.crewai.com/t/llm-response-error-valueerror-invalid-response-from-llm-call-none-or-empty/4425/1 "2025-03-06T06:07:10Z")
Hey CrewAi team,
Can someone please **prioritize this error**? I’m noticing that multiple users of crewai are also running into the same issue when using open-source LLMs.
Here’s the problem:
✅ Everything works **perfectly** end-to-end when using OpenAI’s GPT models.
❌ But the moment I switch to models like **Qwen, LLaMA 3.1, DeepSeek, or Mistral** , I hit an **LLM response error** …
Error I am getting:
**Received None or empty response from LLM call.**
\*\* An unknown error occurred. Please check the details below.\*\*
\*\* Error details: Invalid response from LLM call - None or empty.\*\*
\*\* An unknown error occurred. Please check the details below.\*\*
\*\* Error details: Invalid response from LLM call - None or empty.\*\*
2025-03-06 11:15:07,059 - ERROR - Error in rag\_fucntion: Invalid response from LLM call - None or empty.
Traceback (most recent call last):
File “/home/sranjan710/miniconda3/envs/py310prod/lib/python3.10/site-packages/crewai/agent.py”, line 248, in execute\_task
result = self.agent\_executor.invoke(
File “/home/sranjan710/miniconda3/envs/py310prod/lib/python3.10/site-packages/crewai/agents/crew\_agent\_executor.py”, line 115, in invoke
raise e
File “/home/sranjan710/miniconda3/envs/py310prod/lib/python3.10/site-packages/crewai/agents/crew\_agent\_executor.py”, line 102, in invoke
formatted\_answer = self.\_invoke\_loop()
File “/home/sranjan710/miniconda3/envs/py310prod/lib/python3.10/site-packages/crewai/agents/crew\_agent\_executor.py”, line 166, in \_invoke\_loop
raise e
File “/home/sranjan710/miniconda3/envs/py310prod/lib/python3.10/site-packages/crewai/agents/crew\_agent\_executor.py”, line 140, in \_invoke\_loop
answer = self.\_get\_llm\_response()
File “/home/sranjan710/miniconda3/envs/py310prod/lib/python3.10/site-packages/crewai/agents/crew\_agent\_executor.py”, line 217, in \_get\_llm\_response
raise ValueError(“Invalid response from LLM call - None or empty.”)
ValueError: Invalid response from LLM call - None or empty.
Please **look into this error on priority**. Right now, it feels pointless to have so many LLM providers if **95% of their models aren’t working with CrewAI Agents**.
Thanks,
Sumit
---
### Author:  [@Dhruvin\_5179](https://community.crewai.com/u/Dhruvin_5179)
#### Post date: [March 6, 2025, 10:35am UTC](https://community.crewai.com/t/llm-response-error-valueerror-invalid-response-from-llm-call-none-or-empty/4425/2 "2025-03-06T10:35:25Z")
Hey CrewAI team,
I am facing the same error as @sranjan719 while using the `Qwen/Qwen2.5-VL-3B-Instruct` model. I am hosting the model on a server using vLLM and setting `api_base` as the API endpoint, but it’s not working. I have tried multiple solutions, but nothing has worked so far.
llm = LLM(
model=“huggingface/Qwen/Qwen2-VL-2B-Instruct”,
api\_base=“[http://0.0.0.0:8000/v1/chat/completions](http://0.0.0.0:8000/v1/chat/completions)”,
)
Error:
Error during LLM call: litellm.BadRequestError: HuggingfaceException - {“object”:“error”,“message”:“[{‘type’: ‘missing’, ‘loc’: (‘body’, ‘messages’), ‘msg’: ‘Field required’, ‘input’: {‘inputs’: ‘\<|im\_start|\>system\nYou are a helpful assistant.\<|im\_end|\>\n\<|im\_start|\>user\nYou are Image Analyst. Expert in visual analysis with deep knowledge of design, composition, objects, patterns, and features. Can accurately describe and interpret images across various contexts.\nYour personal goal is: Analyze the given image and provide detailed insights based on the provided question\nTo give my best complete final answer to the task respond using the exact following format:\n\nThought: I now can give a great answer\nFinal Answer: Your final answer must be the great and the most complete as possible, it must be outcome described.\n\nI MUST use these formats, my job depends on it!\n\nCurrent Task: Analyze the image located at [https://as1.ftcdn.net/v2/jpg/10/64/66/34/1000\_F\_1064663493\_PG2uY9VYvZPVxXvmFwIOvGDSNlfMDeIL.jpg](https://as1.ftcdn.net/v2/jpg/10/64/66/34/1000_F_1064663493_PG2uY9VYvZPVxXvmFwIOvGDSNlfMDeIL.jpg) and describe the word written on the image.\n\nThis is the expect criteria for your final answer: A detailed analysis and response based on the given image and question.\nyou MUST return the actual complete content as the final answer, not a summary.\n\nBegin! This is VERY important to you, use the tools available and give your best Final Answer, your job depends on it!\n\nThought:\<|im\_end|\>\n\<|im\_start|\>assistant\n’, ‘parameters’: {‘stream’: False, ‘stop’: [‘’, ‘\nObservation:’], ‘details’: True, ‘return\_full\_text’: False}, ‘stream’: False}}, {‘type’: ‘missing’, ‘loc’: (‘body’, ‘model’), ‘msg’: ‘Field required’, ‘input’: {‘inputs’: ‘\<|im\_start|\>system\nYou are a helpful assistant.\<|im\_end|\>\n\<|im\_start|\>user\nYou are Image Analyst. Expert in visual analysis with deep knowledge of design, composition, objects, patterns, and features. Can accurately describe and interpret images across various contexts.\nYour personal goal is: Analyze the given image and provide detailed insights based on the provided question\nTo give my best complete final answer to the task respond using the exact following format:\n\nThought: I now can give a great answer\nFinal Answer: Your final answer must be the great and the most complete as possible, it must be outcome described.\n\nI MUST use these formats, my job depends on it!\n\nCurrent Task: Analyze the image located at [https://as1.ftcdn.net/v2/jpg/10/64/66/34/1000\_F\_1064663493\_PG2uY9VYvZPVxXvmFwIOvGDSNlfMDeIL.jpg](https://as1.ftcdn.net/v2/jpg/10/64/66/34/1000_F_1064663493_PG2uY9VYvZPVxXvmFwIOvGDSNlfMDeIL.jpg) and describe the word written on the image.\n\nThis is the expect criteria for your final answer: A detailed analysis and response based on the given image and question.\nyou MUST return the actual complete content as the final answer, not a summary.\n\nBegin! This is VERY important to you, use the tools available and give your best Final Answer, your job depends on it!\n\nThought:\<|im\_end|\>\n\<|im\_start|\>assistant\n’, ‘parameters’: {‘stream’: False, ‘stop’: [‘’, ‘\nObservation:’], ‘details’: True, ‘return\_full\_text’: False}, ‘stream’: False}}]”,“type”:“BadRequestError”,“param”:null,“code”:400}
Does Qwen support CrewAI/LiteLLM?
---
### Author:  [@Ashis](https://community.crewai.com/u/Ashis)
#### Post date: [March 12, 2025, 7:16am UTC](https://community.crewai.com/t/llm-response-error-valueerror-invalid-response-from-llm-call-none-or-empty/4425/3 "2025-03-12T07:16:56Z")
Hello All,
Same error for me as well , seems like changing the max\_iter parameter while constructing agent to a lower value some how fixing the error , it seems the error is specific to some models , i was getting error for “groq/llama-3.1-8b-instant” when i changed to a higher model “groq/llama-3.3-70b-specdec” the same code worked fine. Intermittent issue and changing max\_iter param and switching models some how working for me.
---
### Author:  [@maxmoura](https://community.crewai.com/u/maxmoura)
#### Post date: [March 12, 2025, 11:52pm UTC](https://community.crewai.com/t/llm-response-error-valueerror-invalid-response-from-llm-call-none-or-empty/4425/4 "2025-03-12T23:52:12Z")
Are you sure it’s not a naming issue, guys?
As you know, CrewAI uses the LiteLLM library for connections with LLMs, right? And the [LiteLLM documentation about VLLM](https://docs.litellm.ai/docs/providers/vllm) says that the `model` parameter should be `model="hosted_vllm/"`. So, in CrewAI, it would be something like this:
```python
from crewai import LLM
llm = LLM(
model='hosted_vllm/',
api_base='your-hosted-vllm-server',
)
```
---
### Author:  [@Dhruvin\_5179](https://community.crewai.com/u/Dhruvin_5179)
#### Post date: [March 13, 2025, 5:22am UTC](https://community.crewai.com/t/llm-response-error-valueerror-invalid-response-from-llm-call-none-or-empty/4425/5 "2025-03-13T05:22:05Z")
I already tried this, @maxmoura, but it’s still not working. I think it’s a template-related issue. I tried it with the Qwen2.5-VL-3B-Instruct model.
---
### Author:  [@system](https://community.crewai.com/u/system)
#### Post date: [April 5, 2025, 6:08am UTC](https://community.crewai.com/t/llm-response-error-valueerror-invalid-response-from-llm-call-none-or-empty/4425/6 "2025-04-05T06:08:00Z")
This topic was automatically closed after 30 days. New replies are no longer allowed.