the error is

curl: (35) OpenSSL SSL_connect: Connection reset by peer in connection to chatgpt-api.xxx.workers.dev:443

I use vscode plugin: https://github.com/gencay/vscode-chatgpt

I configured the following proxy in settings.json:

{
    "chatgpt.gpt3.apiBaseUrl": "https://chatgpt-api.shn.hk"
}

But report an error:

Your method: 'GPT3 OpenAI API Key' and your model: 'gpt-3.5-turbo' may be incompatible or you may have exhausted your ChatGPT subscription allowance. (HTTP 404 Not Found) ChatGPT error 404: Not found

比如 https://api.openai.com/v1/chat/completions 和 https://chatgpt-api.shn.hk/v1/chat/completions 对齐。因为有很多其他开发者开发的应用中很多都提供了自定义域名功能，比如我填 https://xxx.com 它们软件底层就会自动处理成 https://xxx.com/v1/chat/completions ，但是用本项目提供的域名的话，填 https://chatgpt-api.shn.hk 它们会处理成 https://chatgpt-api.shn.hk/v1/chat/completions ，那么就无法使用到本项目的功能了。

以及， https://chatgpt-api.shn.hk 也可能随时被墙

Can we deploy this on cloudflare?

Please fix this
OpenAI API responded: { "error": { "message": "Your access was terminated due to violation of our policies, please check your email for more information. If you believe this is in error and would like to appeal, please contact [email protected].", "type": "access_terminated", "param": null, "code": null } }

As I am a personal developer, Open AI recharge is not supported in my area, and my personal free API account has been banned recently, so this project is of great help to me! I hope it can survive for a long time. As a student, I don't have a lot of money, but I have contributed some small money to you, hoping to help this project. If you can't stick to this free project, I hope I can pay for it. Please let me know before you finish this project. Thank you. Sincerely Ming-XF.

Description: I am unable to access the Free ChatGPT website at https://freechatgpt.chat/. The page is taking too long to load and seems unresponsive. I have tried accessing it multiple times, but the issue persists. I suspect it's not a network issue on my end.

Steps to reproduce:

1. Visit https://freechatgpt.chat/

2. Try to connect with valid credentials

3. Observe the slow loading and unresponsiveness of the page

I have tried accessing the website from different devices and networks, but the issue remains the same. Kindly assist in resolving this issue.

Thank you.

Hey, I am getting this error:

`OpenAI API responded:

{
"error": {
"message": "You exceeded your current quota, please check your plan and billing details.",
"type": "insufficient_quota",
"param": null,
"code": null
}
}
`

这个chatgpt的api是一个字一个字返回的吗还是一下返回全部文本信息

2023-03-16T15:19:39.291Z # trying to hit the 8080 port using http

2023-03-16T15:19:42.376Z # it looks that no process is listening to the 8080 port using http

2023-03-16T15:19:42.377Z # app did not turn healthy after several checks

2023-03-16T15:19:42.378Z # deployment failed

Does anyone know how I can fix it?

I have a lot to donate if that's open.

I wondered, if you want to host your own api, do you have to pay for the tokens?

The LLaMA model has been leaked and many people are running it locally on their machine. Maybe we can collectively host this and create an endpoint to use when the chatgpt endpoint gets restricted or saturated
here are two projects to look at from @cocktailpeanut and @ggerganov
https://github.com/cocktailpeanut/dalai
https://github.com/ggerganov/llama.cpp

Does this api support GPT-4 ?

Al application developers do not need to develop a power grid first in order for users to use electronic appliances.

Hello, Ms. Ayaka!

I received an error message that says "KV PUT failed: 400 Invalid expiration of 1678125600. Expiration times must be at least 60 seconds in the future," while using your unmodified code. Based on the error message, it seems that the expiration time used in the kv.put function is not valid because the expiration parameter provided is less than 60 seconds ahead of the current time.

To resolve the error, we can update the expiration time by adding 60 seconds to the current time using the moment library's add() method before converting it to a Unix timestamp using the unix() method. We can achieve this by using the following code:

const rateLimitExpiration = utcNow.startOf('hour').add(1, 'hour').add(60, 'seconds').unix();

By doing this, the expiration parameter passed to the kv.put function will always be at least 60 seconds in the future, and the error should no longer occur.

Please let me know your thoughts on this solution.

Hi Ma'am Ayaka!

Thank you for this. Hoping you can help me how to set up the wrangler.toml properly since I have multiple Cloudflare accounts connected to my main CF account.

Let's say I wanted to use the AAA account.

Thank you :)

✘ [ERROR] More than one account available but unable to select one in non-interactive mode.

If you think this is a bug then please create an issue at https://github.com/cloudflare/workers-sdk/issues/new/choose
Please set the appropriate account_id in your wrangler.toml file.
Available accounts are (<name>: <account_id>):
AAA: ece12a4fbc7b131fc17d118
BBB: c1b0fa2f2e392fd5dbbca5e3e
CCC: 04ecf3126ab6a336d11c[25](https://github.com/xxxxx/actions/runs/43xxjobs/756xx756#step:5:26)7d4c746116
DDD: f618f8ef0557ef4d9c898c6c[26](https://github.com/xxxxx/actions/runs/43xx0380/jobs/7xx56#step:5:27)7821a
EEE: 5f1d81df6a7ba9461f756a8faeb34[29](https://github.com/xxxxx/actions/runs/4xxx0/jobs/75xx56#step:5:30)

Hi Ayaka!

I am writing to suggest improving the reliability of the ChatGPTAPIFree by implementing proper error handling for MongoDB connections.

Currently, the code assumes that the connection to MongoDB will always succeed, but there are several reasons why the connection could fail, such as network issues or invalid credentials. As a result, the lack of proper error handling may lead to critical issues, such as data loss or incorrect data processing.

I recommend adding error handling for MongoDB connection failures to address this issue. By doing so, you can improve the reliability of the API and reduce the risk of data loss or incorrect data processing.

Thank you for considering my suggestion.

Sample code:

• The MongoClient is created outside the app variable and connected to MongoDB using the connect() method. If an error occurs, the code logs the error and exits the process with an exit code of 1 to indicate a fatal error.
• The MongoDB collection is created, and the indexes are created as before.
• The catch block of the /v1/ route now logs the error to the console before sending a generic error message back to the client.
• This implementation should now handle errors related to MongoDB connections.

const client = new MongoClient(uri);
client.connect((err) => {
  if (err) {
    console.error(err);
    process.exit(1);
  } else {
    console.log('Connected successfully to MongoDB at ' + uri);
  }
});

    resUpstream.body.pipe(res);
  } catch (error) {
    console.error(error);
    res.status(500).set(corsHeaders).type('text/plain').send('Internal server error');
  }
});

Getting this when making a request:

HTTP 429

OpenAI API responded with:

{
    "error": {
        "message": "You exceeded your current quota, please check your plan and billing details.",
        "type": "insufficient_quota",
        "param": null,
        "code": null
    }
}

it seems like the rate limit has been reached

Hello, I send a request with python, but I got this error.

requests.exceptions.JSONDecodeError: [Errno Expecting value] <!doctype html><title>429</title>429 Too Many Requests: 0

Bests Regards

Hello Ms. Ayaka,

I successfully setup my own instance unfortunately I am having a problem with CORS policy, here are the error messages I am getting from Google Run and web console:

{
httpRequest: {11}
insertId: "640c4eXXXXXXXXXbef81ca15"
labels: {5}
logName: "projects/XXXXX/logs/run.googleapis.com%2Frequests"
receiveTimestamp: "2023-03-11T09:50:XXXXZ"
resource: {2}
severity: "ERROR"
spanId: "4361714022011534635"
textPayload: "The request failed because either the HTTP response was malformed or connection to the instance had an error. Additional troubleshooting documentation can be found at: https://cloud.google.com/run/docs/troubleshooting#malformed-response-or-connection-error"
timestamp: "2023-03-11T09:50:3XXXZ"
trace: "pXXXX14faaa"
}

Access to fetch at 'https://XXXXX.a.run.app/v1/' from origin 'https://aXXXXX.com' has been blocked by CORS policy: No 'Access-Control-Allow-Origin' header is present on the requested resource. If an opaque response serves your needs, set the request's mode to 'no-cors' to fetch the resource with CORS disabled.

Response Headers

access-control-allow-headers: Content-Type, Authorization
access-control-allow-methods: GET, HEAD, POST, OPTIONS
access-control-allow-origin: *
alt-svc: h3=":443"; ma=xx29=":443"; ma=2592000,h3-Q0xxx:443"; ma=259xx0,quic=":443"; ma=2592000; v="46,43"
content-length: 9
content-type: text/plain; charset=utf-8
date: Sat, 11 Mar 2023 09:59:10 GMT
server: Google Frontend
x-cloud-trace-context: ef8663xxx533eed

Thank you.

Your access was terminated due to violation of our policies, please check your email for more information. If you believe this is in error and would like to appeal, please contact [email protected].

The above error occurs when I use。

It doesn't work and report error.

Can we use Upstash Redis for rate limiting requests instead of hardcoding it on the worker code?

Link: https://upstash.com/
Github: https://github.com/upstash/upstash-redis
Cloudflare Workers Example: https://github.com/upstash/upstash-redis/tree/main/examples/cloudflare-workers

The error is getting too frequent.
Thanks for making chatGPT free to developers.

Just saw this amazing initiative. I really wanted to try out ChatGPT, but I didn't have a Phone Number. Now I can experiment with it. Thanks very much.

I do want to support this project, but I can't financially, sorry. But I am a student learning developing/programming, and probably can help with anything you want.

Just tell me 😉

can you do the same for dalle-2?

I created a docker image and ran it using:

docker build -t chatgpt:1.0 .
docker run -d -p 8080:8080 chatgpt:1.0
I exposed port 8080 of the container, so I could use it on my localhost browser.

The server is listening on 8080 but when I tried to run it, I get a 'not found' error.

index.js references some /v1/chat/completions folder but I don't see it in the local directory.

I renamed ".env.example" to ".env" and entered two OpenAI API keys there. Prior to renaming it, I was getting some other error related to the JSON parse error.

One error is leading to fixing another thing and now I am stuck on this 'Not found' page.

Any light on this matter would be greatly appreciated. Thanks.

PS: I am running Docker Desktop latest version, x86 machine, Windows 10.

Hi Ayaka,

I want to self-host my own API endpoint, please kindly share what Google Cloud Armor rules did you use?

Thank you.

Because you are literally paying for others without any rewards. I am thankful about it, but just concerning: are you okay?

Free with limited quota? I think so this is not free, just trial.

OpenAI API responded:

{
    "error": {
        "message": "You exceeded your current quota, please check your plan and billing details.",
        "type": "insufficient_quota",
        "param": null,
        "code": null
    }
}