#Single chat session limit

1 messages · Page 1 of 1 (latest)

feral brook
#

Is there a limit to how long a single chat session can go, I've twice now run into a point where a single chat session has started glitching

errant yarrowBOT
#

Hi @feral brook. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:

#

It seems like you might be experiencing issues related to session limits. Based on the information available, there are a couple of relevant points regarding session limits:

  1. Realtime API Session Limit: If you are using the Realtime API, there is a 15-minute limit for session time for WebSocket connections. After this limit, the server will disconnect. This time refers to the wallclock time of the session connection, not the length of input or output audio. Learn more here (https://platform.openai.com/docs/guides/realtime#handling-long-conversations).

  2. Token Limits: Many APIs, including those from OpenAI, have token limits for both input and output. For example, GPT-3.5 has a limit of 4,096 tokens, and GPT-4 can have limits of 8,192 or 32,768 tokens, depending on the version. If you exceed these limits, the assistant may ignore additional messages or responses. Learn more here (https://community.openai.com/t/1018601).

If your session is glitching, it might be due to reaching these limits. You can try reducing the number of tokens by summarizing messages or batching requests to manage the session more effectively. If you need further assistance, you might want to check the API documentation or logs for more specific details on the limits you are encountering.

Note: you can ask me a follow up question by @ mentioning me again :speech_balloon:

#

@feral brook Kapa.ai is still learning and improving, please let me know how I did by reacting below.