Hey there, just started using AgentGPT today, and any time I make an agent, if I chat with the bot after the first process is done I get this error.

This model's maximum context length is 4096 tokens. However, your messages resulted in 5488 tokens. Please reduce the length of the messages.

Am I doing something wrong or is this a bug?

can i use it with local llm?

Gift 50$ - [steamcommunity.com/gift/50](https://bit.ly/4diEcJT ) @everyone @here

Hey @Ali Ghanem Hibbant how do I get passed the cap on tasks? I can seem to finish making a report..

Can I pay?

how do I limit the amount of loops it does?

@Nic | Antisuns É

Hi, I'm technically not really moderating this forum any more. Please talk to the @User

ok thanks @Adam any luck here?

I can't seem to ever get a finished tasking done.

it always goes past the 26 cycles

also how to I get a human in the loop interaction for adding input to tasks?

@Jay B Hey sorry for the really late reply, I just noticed the ping from you - it's my fault - and honestly I'm so busy with something else. What did you mean get passed the cap of the tasks?

oh hey!

I mean it keep s going passed the 26 loops

Is there any plans on developing the AgentGPT or it is a abandoned project at this moment? There were no single accepted PR in its repo in the past 5 months. If so, then what alternatives would you suggest to try? https://cdn.discordapp.com/attachments/1110451216946835466/1259763839910543420/image.png?ex=668cde00&is=668b8c80&hm=429e88de853353280724370fa12dff4dd53776ef67f57f89c52f94d0834e9e8b&

💯

Am I doing something wrong? I thought one of the things that this system apparently does well is web scraping. I asked it to provide contact details for some companies from the internet. It then says it cannot do this and then tells me how to do this myself! A free python script can do web scraping (not always very well). Thanks. Sanwar

Is the current version of agentGPT (self-hosted and all that) able to utilize chatGPT 4.o and/or 4.o-mini yet? If so, I might have a bit of an issue seeing as in the "advanced settings" under "model" I can only find/select "gpt-3.5-turbo"

Not had a response to multiple messages on Discord and via email. Not very good at web scraping. Better off using a php script for free. I will ask my bank for a refund.

{user} here is a gift from valorant cloveses.pro - go enter promo code - h574zx @everyone

Hi guys I was wondering if there is an easy way for me to install AgentGPT on Vercel and host this online ratker than through Docker?

Does local AgentGPT allow utilizing a locally-running LLM from Hugging Face?

yes if you point your api endpoints there and its openai compatible

i just purchased AgentGPT Pro plan which includes "Unlimited web search capabilities". Every prompt I try responds that as an AI agent, it cannot access the web. How do I adjust my prompt or settings to get it to use web search?

hey, how I can reach out to discuss a business inquiry?

yeah

im wondering this too

Howdy, all! Say - I've been successfully using my OpenAI API Key, running AgentGPT on GPT 3.5 turbo16k without issue. Wanted to try a specific prompt with gpt-4 but... any ideas how/why 'The model gpt-4 does not exist or you do not have access to it‽ https://cdn.discordapp.com/attachments/1110451216946835466/1346947061425111231/image.png?ex=67ca09b4&is=67c8b834&hm=dd9be406b7e97f7920128a28522303331ed4bd442394768287f7d73ac98353f0&