After conducting multiple conversations in the same session, my question tokens keep exceeding the limit

General
, ,
1

After conducting multiple conversations in the same session, my question tokens keep exceeding the limit. Is there something in the Agent that is increasing the token count of my questions? I am using the knowledge base and memory.

MyAgent：

 self.agent = Agent(
            #模型
            model= model,
            #用户id
            user_id= role.roleid,
            #介绍
            introduction=dedent("""
            You are a professional TRPG game player.
            You excel at analyzing and using tool functions, knowledge bases, past memories, mastering game rules and game situations.
            You are skilled at role-playing specified characters, and planning character actions based on known information to make reasonable decisions.
            """),
            #重复对话
            session_id=self.session_id,

            #记忆引用
            memory=self.memory,
            enable_agentic_memory=self.is_memory,
            enable_user_memories=self.is_memory,
            add_memory_references=self.is_memory,
            add_session_summary_references=self.is_memory,
            enable_session_summaries=self.is_memory,
            #工具
            tools = tools,
            show_tool_calls=True,
            tool_choice=tool_call,
            #知识
            knowledge=knowledge,
            search_knowledge=self.is_knowledge
        )
2

Hi @komiya , thanks for reaching out and supporting Agno. I’ve shared this with the team, we’re working through all requests one by one and will get back to you soon.
In the meantime please refer the FAQ: Tokens-per-minute rate limiting - Agno

3

I want to know whether adding memory will increase the token for each query.

4

Hey @komiya

Our memory features (like chat history, user memories, and session summaries) are designed to help agents feel more intelligent and personal. But as you noticed, this richer context does mean more tokens are used as conversations get longer. It’s a balance: the more your agent “remembers,” the more resources it needs.

You can have a look at our Memory docs: Memory - Agno