Issue: Unable to use Volcengine, SiliconFlow, and 302AI.
SiliconFlow: My previous workflow was working fine, but now the model list is completely empty (no models are displaying).
Volcengine: The models show up in the list, but I receive an error upon calling them.
Note: I have already verified that I have enabled the necessary permissions for the corresponding models.
My version is 2.6.3
Hello @xincheng thanks for reaching out.
You need to review your account on langchain to know if you have access to that model.
The HTTP Error indicates the model you’re requesting does not exist or you no longer have access to it. The Model or endpoint doubao-embedding-text-240715 does not exist or you do not have access to it.
Otherwise - if you’ve checked and you’re certain the issue is not langchain. I would recommend remove this node completely and restart your credentials - see if a complete node recreation will resolve any caching problems.
If you’re still having the issue please share your canvas code - I’ll replicate on my side and see if I have same problem.
On Stand By
{
“nodes”: [
{
“parameters”: {},
“type”: “n8n-nodes-base.manualTrigger”,
“typeVersion”: 1,
“position”: [
1936,
944
],
“id”: “e9937822-5295-45d7-a79d-eba1eae47c25”,
“name”: “When clicking ‘Execute workflow’”
},
{
“parameters”: {
“fileSelector”: “/images/测试-心橙 AI 工作室：以智能之橙，点亮产业创新之心.pdf”,
“options”: {}
},
“type”: “n8n-nodes-base.readWriteFile”,
“typeVersion”: 1,
“position”: [
2144,
944
],
“id”: “7255effb-07a6-4fb0-a6c3-970626389fff”,
“name”: “Read/Write Files from Disk”
},
{
“parameters”: {
“mode”: “insert”,
“pineconeIndex”: {
“__rl”: true,
“value”: “n8n-1208”,
“mode”: “list”,
“cachedResultName”: “n8n-1208”
},
“options”: {
“pineconeNamespace”: “ceshi”
}
},
“type”: “@n8n/n8n-nodes-langchain.vectorStorePinecone”,
“typeVersion”: 1.3,
“position”: [
2352,
944
],
“id”: “3987c8ef-3892-4583-afa3-81cb22346514”,
“name”: “Pinecone Vector Store”,
“credentials”: {
“pineconeApi”: {
“id”: “KE0J7XDddzVy8q2b”,
“name”: “PineconeApi account”
}
}
},
{
“parameters”: {
“dataType”: “binary”,
“options”: {}
},
“type”: “@n8n/n8n-nodes-langchain.documentDefaultDataLoader”,
“typeVersion”: 1.1,
“position”: [
2480,
1184
],
“id”: “b3d4ccfe-a208-4f71-b1d5-76817d2fb058”,
“name”: “Default Data Loader”
},
{
“parameters”: {
“model”: “doubao-embedding-text-240715”,
“options”: {}
},
“type”: “@n8n/n8n-nodes-langchain.embeddingsOpenAi”,
“typeVersion”: 1.2,
“position”: [
2352,
1184
],
“id”: “370ac084-23af-4a33-ab3d-280bbaf0f985”,
“name”: “Embeddings OpenAI2”,
“credentials”: {
“openAiApi”: {
“id”: “vW7oXv8cUvtDfYEw”,
“name”: “302”
}
}
}
],
“connections”: {
“When clicking ‘Execute workflow’”: {
“main”: [
[
{
“node”: “Read/Write Files from Disk”,
“type”: “main”,
“index”: 0
}
]
]
},
“Read/Write Files from Disk”: {
“main”: [
[
{
“node”: “Pinecone Vector Store”,
“type”: “main”,
“index”: 0
}
]
]
},
“Default Data Loader”: {
“ai_document”: [
[
{
“node”: “Pinecone Vector Store”,
“type”: “ai_document”,
“index”: 0
}
]
]
},
“Embeddings OpenAI2”: {
“ai_embedding”: [
[
{
“node”: “Pinecone Vector Store”,
“type”: “ai_embedding”,
“index”: 0
}
]
]
}
},
“pinData”: {},
“meta”: {
“templateCredsSetupCompleted”: true,
“instanceId”: “72b9b14682050c5676171cb5e88adfd1af09dd60cb9447df66d18ba4072efb4e”
}
}
I just tried version 1.2.X where I originally created this workflow, and it still works perfectly fine when using it directly in that version.
ceshi.json (3.4 KB)
Version 1.121.2
Hi @xincheng Welcome!
I see the issue try switching the model from the current one to
@xincheng set your BASE URL in the node credentials to this https://api.openai.com and then save and you will be able to see text-embeddings models and then choose the large one. And that would start working.
Since we have difficulty purchasing OpenAI’s official API, we are using a compatible alternative that follows the OpenAI protocol. We can use it just like the original API by simply replacing the Base URL and API Key.
@xincheng According to the error which clearly says that MODEL endpoint <model_name> does not exist, which means you have to use official API key and base URL which i have suggested to be able access the standard models which are text-embedding, OR see which models actually exist in that BASE URL model provider alternative you are using and then use that although as it seems like they only support 3-4 models and neither those are working, so possible and best approach would be to use official OpenAI base url with their key.
Solution Found:
The trailing slash at the end of the Base URL must be removed.
Change it to:
https://api.siliconflow.cn/v1 (no / at the end).
nice work - passed us completely haha. I was more stuck that you’d indicated that it had worked prior and something seemed to have changed. glad it’s working for you now.
In n8n version 2.6.3, the issue where the OpenAI Embedding node fails to fetch the model list is typically caused by stricter Credential validation or Base URL path checks in the new version.
Based on community feedback and version differences, it is recommended to follow these steps to troubleshoot:
The OpenAI node in the new version (2.x) is highly sensitive to URL formatting.
Problem Description: If you entered https://api.openai.com/v1/ (with a trailing slash) in the credentials, the Chat node might still function, but the Embedding node may fail to fetch the model list due to an extra slash being added during path concatenation.
Solution: Go to Credentials, remove the trailing slash at the end of the Base URL, and change it to: https://api.openai.com/v1
