diff --git a/src/libs/agent-runtime/huggingface/index.ts b/src/libs/agent-runtime/huggingface/index.ts
index b31a12b1af27..552652f9470a 100644
--- a/src/libs/agent-runtime/huggingface/index.ts
+++ b/src/libs/agent-runtime/huggingface/index.ts
@@ -1,4 +1,5 @@
 import { HfInference } from '@huggingface/inference';
+import urlJoin from 'url-join';
 
 import { AgentRuntimeErrorType } from '../error';
 import { ModelProvider } from '../types';
@@ -23,7 +24,9 @@ export const LobeHuggingFaceAI = LobeOpenAICompatibleFactory({
     createChatCompletionStream: (client: HfInference, payload, instance) => {
       const { max_tokens = 4096} = payload;
       const hfRes = client.chatCompletionStream({
-        endpointUrl: instance.baseURL,
+        endpointUrl: instance.baseURL
+          ? urlJoin(instance.baseURL, payload.model)
+          : instance.baseURL,
         max_tokens: max_tokens,
         messages: payload.messages,
         model: payload.model,