Skip to main content

Queue messages

I would like to be able to add followups to my message before the model finishes writing out the code. This may be fixed just by the models being faster. They have been slow for me today, running this on a raspberry pi, but I would like to be able to press enter send the next thing for it to respond to instead of waiting for it to finish to be able to send the next message.

This would kind of be like premoving in chess . com. I can see part of the answer and I have a follow up, I don’t want to have to wait for the model to finish typing for me to be able to ask a followup.

An even better version of this, this is one of my biggest problems with ai chat bots.

When I type I often type several short messages instead of one long message.

I haven’t seen a chat app that supports that yet, especially if it is a thinking model like o3 mini high, or claude 3.7 reasoning it would be nice if it took in my next message as well if it has not started showing tokens yet.

Sometimes it is only after I hit send I realise I need to add something else and I would prefer to be able to quickly type something and push send instead of having to stop the generation and then go up manually copy the previous prompt and then add my additional context or question to the previous prompt.

Status: Gathering Interest6 comments

Log in to comment and vote

Comments6

  • Amaranth Dew

    Yeah, same.

    Codex/Claude Code have trained me into this mental model too much :)

    • Mark Florkowski

      Team

      Can you tell me more about how you use this?

      • Amaranth Dew

        1. When I m see a model staring into a direction I think might need modification, I immediately note down my correction/clarification and would prefer just queuing it (although I would still like it to complete the response).

        2. I have a follow-up thought that I really need to put down and send because it would also let me accomplish the next part of the task.

        Having to wait until after the model generates its response(s) makes me tied to the inference speed of the model, which is provider dependent.

  • Indigo Stew

    would be really cool to have this feature..
    it would really be very helpful, like to queue like 3-5messages..

  • Blush Sand

    •

    Merged request

    Brainstorming Notes / Messages

    Oftentimes, I want to brainstorm something, and want to have an ai generate a rough idea of the code I want to write. It would be helpful to send multiple messages without immediately prompting the chat bot for a response, essentially just to take notes from various sources, and after entering a few, then prompt the bot for a response.

    TLDR: send multiple user messages before the bot responds

  • Purple Cookie

    •

    Merged request

    Queue messages

    As chat 1 is generating, change the behavior such that sending a second chat waits for this to be finished and then continues the convo from there instead of overriding (similar to cursor)