I am getting this error:
Problem in node ‘PartialExecutionToolExecutor‘ Error executing tool: Bad Request
Mine is local instance. I am trying to connect it to a collection in my qdrant vector db so for each chat/query it should search in qdrant for matching embedding and return the response.
hi @Deepak_Itkar !
What is the vector dimension of the Qdrant collection?
Which embedding model is configured in the Embeddings Ollama node, and what is its output dimension?
Does the Qdrant collection use a custom vector name, and if so, is the same name configured in the Qdrant Vector Store node?
Based on the current setup and the reported error (Bad Request from Qdrant during AI Agent tool execution), the most likely cause is a mismatch between the embedding dimension or the vector name used by Ollama and the configuration of the Qdrant collection.
vector dimension = 1024
embedding model = nomic-embed-text
Does the Qdrant collection use a custom vector name, and if so, is the same name configured in the Qdrant Vector Store node? I see the list of collections under “Qdrant Collection” and selected the desired one.
The issue is caused by a vector name mismatch.
Your Qdrant collection uses the vector name default, while n8n queries vector by default, so you must explicitly set Vector Name = default in the Qdrant Vector Store node.