From f06e3d3efac74aa4cc7bc133e2bb3abe5008a3d4 Mon Sep 17 00:00:00 2001
From: Alfons <alfonsxh@gmail.com>
Date: Wed, 5 Nov 2025 15:52:17 +0800
Subject: [PATCH] fix: disabling potential `thinking` param for model testing
 (#1733)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

* fix: 禁用模型默认思考功能以减少测试延迟

- 调整导入语句顺序
- 为没有显式设置 thinking 参数的模型添加禁用配置
- 避免某些模型厂商默认开启思考功能导致的测试延迟

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

* fix: 确保 extra_args 为空时也禁用思考功能

修复条件判断逻辑，当 extra_args 为空字典时也会添加思考功能禁用配置

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

* perf(fe): increase default timeout

* perf: llm model testing prompt

---------

Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Junyan Qin <rockchinq@gmail.com>
---
 pkg/api/http/service/model.py            | 12 +++++++++---
 web/src/app/infra/http/BaseHttpClient.ts |  2 +-
 2 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/pkg/api/http/service/model.py b/pkg/api/http/service/model.py
index 036c1b9c..17297ed1 100644
--- a/pkg/api/http/service/model.py
+++ b/pkg/api/http/service/model.py
@@ -1,13 +1,14 @@
 from __future__ import annotations
 
 import uuid
+
 import sqlalchemy
+from langbot_plugin.api.entities.builtin.provider import message as provider_message
 
 from ....core import app
 from ....entity.persistence import model as persistence_model
 from ....entity.persistence import pipeline as persistence_pipeline
 from ....provider.modelmgr import requester as model_requester
-from langbot_plugin.api.entities.builtin.provider import message as provider_message
 
 
 class LLMModelsService:
@@ -104,12 +105,17 @@ class LLMModelsService:
         else:
             runtime_llm_model = await self.ap.model_mgr.init_runtime_llm_model(model_data)
 
+        # 有些模型厂商默认开启了思考功能，测试容易延迟
+        extra_args = model_data.get('extra_args', {})
+        if not extra_args or 'thinking' not in extra_args:
+            extra_args['thinking'] = {'type': 'disabled'}
+
         await runtime_llm_model.requester.invoke_llm(
             query=None,
             model=runtime_llm_model,
-            messages=[provider_message.Message(role='user', content='Hello, world!')],
+            messages=[provider_message.Message(role='user', content='Hello, world! Please just reply a "Hello".')],
             funcs=[],
-            extra_args=model_data.get('extra_args', {}),
+            extra_args=extra_args,
         )
 
 
diff --git a/web/src/app/infra/http/BaseHttpClient.ts b/web/src/app/infra/http/BaseHttpClient.ts
index 019a54e6..cc2c31e4 100644
--- a/web/src/app/infra/http/BaseHttpClient.ts
+++ b/web/src/app/infra/http/BaseHttpClient.ts
@@ -38,7 +38,7 @@ export abstract class BaseHttpClient {
 
     this.instance = axios.create({
       baseURL: baseURL,
-      timeout: 15000,
+      timeout: 30000,
       headers: {
         'Content-Type': 'application/json',
       },