diff --git a/metagpt/provider/base_llm.py b/metagpt/provider/base_llm.py
index 4d00adbc7..52dd96b1a 100644
--- a/metagpt/provider/base_llm.py
+++ b/metagpt/provider/base_llm.py
@@ -43,7 +43,7 @@ class BaseLLM(ABC):
         if system_msgs:
             message = self._system_msgs(system_msgs)
         else:
-            message = [self._default_system_msg()]
+            message = [self._default_system_msg()] if self.use_system_prompt else []
         if format_msgs:
             message.extend(format_msgs)
         message.append(self._user_msg(msg))
diff --git a/metagpt/provider/fireworks_api.py b/metagpt/provider/fireworks_api.py
index 638b0703d..f0af68818 100644
--- a/metagpt/provider/fireworks_api.py
+++ b/metagpt/provider/fireworks_api.py
@@ -64,8 +64,9 @@ class FireworksCostManager(CostManager):
         token_costs = self.model_grade_token_costs(model)
         cost = (prompt_tokens * token_costs["prompt"] + completion_tokens * token_costs["completion"]) / 1000000
         self.total_cost += cost
+        max_budget = CONFIG.max_budget if CONFIG.max_budget else CONFIG.cost_manager.max_budget
         logger.info(
-            f"Total running cost: ${self.total_cost:.4f} | Max budget: ${CONFIG.max_budget:.3f} | "
+            f"Total running cost: ${self.total_cost:.4f} | Max budget: ${max_budget:.3f} | "
             f"Current cost: ${cost:.4f}, prompt_tokens: {prompt_tokens}, completion_tokens: {completion_tokens}"
         )
         CONFIG.total_cost = self.total_cost
diff --git a/metagpt/provider/ollama_api.py b/metagpt/provider/ollama_api.py
index 95b944bf3..8ee04de7d 100644
--- a/metagpt/provider/ollama_api.py
+++ b/metagpt/provider/ollama_api.py
@@ -30,9 +30,9 @@ class OllamaCostManager(CostManager):
         """
         self.total_prompt_tokens += prompt_tokens
         self.total_completion_tokens += completion_tokens
-
+        max_budget = CONFIG.max_budget if CONFIG.max_budget else CONFIG.cost_manager.max_budget
         logger.info(
-            f"Max budget: ${CONFIG.max_budget:.3f} | "
+            f"Max budget: ${max_budget:.3f} | "
             f"prompt_tokens: {prompt_tokens}, completion_tokens: {completion_tokens}"
         )
         CONFIG.total_cost = self.total_cost
diff --git a/metagpt/provider/open_llm_api.py b/metagpt/provider/open_llm_api.py
index 7f5870702..b0c484f5a 100644
--- a/metagpt/provider/open_llm_api.py
+++ b/metagpt/provider/open_llm_api.py
@@ -26,9 +26,9 @@ class OpenLLMCostManager(CostManager):
         """
         self.total_prompt_tokens += prompt_tokens
         self.total_completion_tokens += completion_tokens
-
+        max_budget = CONFIG.max_budget if CONFIG.max_budget else CONFIG.cost_manager.max_budget
         logger.info(
-            f"Max budget: ${CONFIG.max_budget:.3f} | reference "
+            f"Max budget: ${max_budget:.3f} | reference "
             f"prompt_tokens: {prompt_tokens}, completion_tokens: {completion_tokens}"
         )
 
diff --git a/metagpt/provider/zhipuai/zhipu_model_api.py b/metagpt/provider/zhipuai/zhipu_model_api.py
index 72be0f333..c2c1bd3d8 100644
--- a/metagpt/provider/zhipuai/zhipu_model_api.py
+++ b/metagpt/provider/zhipuai/zhipu_model_api.py
@@ -2,6 +2,7 @@
 # -*- coding: utf-8 -*-
 # @Desc   : zhipu model api to support sync & async for invoke & sse_invoke
 
+import json
 import zhipuai
 from zhipuai.model_api.api import InvokeType, ModelAPI
 from zhipuai.utils.http_client import headers as zhipuai_default_headers
@@ -51,7 +52,6 @@ class ZhiPuModelAPI(ModelAPI):
             params=kwargs,
             request_timeout=zhipuai.api_timeout_seconds,
         )
-
         return result
 
     @classmethod
@@ -61,6 +61,8 @@ class ZhiPuModelAPI(ModelAPI):
         resp = await cls.arequest(
             invoke_type=InvokeType.SYNC, stream=False, method="post", headers=headers, kwargs=kwargs
         )
+        resp = resp.decode("utf-8")
+        resp = json.loads(resp)
         return resp
 
     @classmethod
diff --git a/metagpt/provider/zhipuai_api.py b/metagpt/provider/zhipuai_api.py
index addbe58af..865b7fce1 100644
--- a/metagpt/provider/zhipuai_api.py
+++ b/metagpt/provider/zhipuai_api.py
@@ -38,12 +38,11 @@ class ZhiPuAILLM(BaseLLM):
     From now, there is only one model named `chatglm_turbo`
     """
 
-    use_system_prompt: bool = False  # zhipuai has no system prompt when use api
-
     def __init__(self):
         self.__init_zhipuai(CONFIG)
         self.llm = ZhiPuModelAPI
         self.model = "chatglm_turbo"  # so far only one model, just use it
+        self.use_system_prompt: bool = False  # zhipuai has no system prompt when use api
 
     def __init_zhipuai(self, config: CONFIG):
         assert config.zhipuai_api_key
@@ -101,7 +100,6 @@ class ZhiPuAILLM(BaseLLM):
             elif event.event == ZhiPuEvent.ERROR.value or event.event == ZhiPuEvent.INTERRUPTED.value:
                 content = event.data
                 logger.error(f"event error: {content}", end="")
-                collected_content.append([content])
             elif event.event == ZhiPuEvent.FINISH.value:
                 """
                 event.meta