Simple ChatGPT Telegram Multi Chat Bot Windows Console Application
This simple console application interfaces Telegram with ChatGPT, allowing Telegram bots to act as agents relaying messages to and from OpenAI’s ChatGPT, based on a personality assigned to the bot. From Telegram, you can query the bot, which relays the message to the ChatGPT API, queries OpenAI, and receives a response based on the personality of your Telegram bot, as customized within the configuration file. This was a quick one-day project, and I thought I’d share it online. I’ll add more features over time, so check back.
Download:
Download Telegram ChatGPT Bots Windows Console App – 2023-10-17 v0.1a zip
Commands:
Telegram:
create | Creates a chat bot.
edit | Edit a chat bot.
delete | Delete a chat bot.
list | List all chat bots.
ChatGPT:
start | start chatgpt bots.
stop | stop chatgpt bots.
botchat | Two bots talk to each other. Requires two or more bots in same telegram channel and connected to chatgpt.
exit | Exit application.
Getting Started:
To use this application you will need to modify the configuration files to interface the application with your Telegram bots and also your ChatGPT account. Below are the instructions on how to do this.
Telegram Bots:
How to create a bot:
- Start a chat with BotFather.
- Click Menu and select NewBot or type /newbot.
- Enter the bot name.
- Enter the username for the bot, it must end with ‘bot’ and must be unique.
- Congratz, you’ve just created a telegram bot. Copy the HTTP API token into bots.json file. In addition, add your botName (the bot’s telegram username) and full name, the ordinary name of the bot.
- Enter a description of the personality you would like the bot to have. This will guide the bot’s personality when it uses ChatGPT to generate and respond.
Connecting to ChatGPT:
- Visit OpenAI API key page.
- Click “Create new secret key” and copy it to clipboard, then paste it into completionParameters.json apiKey : “your chatgpt api key here” and replace the text in the quotes. IMPORTANT: Never share your API key with anyone as this key can be used to access OpenAI services through your account.
- Configure the settings of completionParameters.json to cater to your needs.
Important: I strongly advise implementing a usage limit for your OpenAI API account. Additionally, it’s advisable to restrict the usage of the API to a private Telegram channel. Querying ChatGPT typically incurs real-world computational costs. You can adjust the “max_tokens” per query using “completionParameters.json,” but please be aware that if your channel is open to the public, other users may also make queries with your bot, potentially leading to additional costs.
Brief explanation of each configuration parameter:
- “apiUrl”: The URL for making requests to the OpenAI ChatGPT API.
- “apiModelsUrl”: The URL for accessing information about available models.
- “apiKey”: Your authentication key to access the ChatGPT API.
- “model”: The specific ChatGPT model to be used, in this case, “text-davinci-003.”
- “prompt”: The initial text or context provided to the model to generate responses.
- “suffix”: An optional additional text to add to the end of generated responses.
- “max_tokens”: The maximum number of tokens in the response, with a range between 1 and 4096.
- “temperature”: A parameter controlling the randomness of the output, with a range between 0.2 and 1.0.
- “top_p”: A value to control the diversity of the output, with a range between 0.0 and 1.0.
- “n”: The number of alternative responses to generate, with a range between 1 and 20.
- “stream”: A boolean indicating whether to stream the response or get it as a whole.
- “logprobs”: Optional setting for obtaining log probabilities for each token in the response.
- “echo”: A boolean indicating whether to include the input prompt in the output.
- “stop”: An optional string that, if encountered in the response, will stop the generation.
- “presence_penalty”: A penalty value to discourage the model from repeating a specific response, with a range between -1.0 and 1.0.
- “frequency_penalty”: A penalty value to discourage the model from using frequent words, with a range between -1.0 and 1.0.
- “best_of”: The number of “n” alternatives to consider and return the best one from, with a range between 1 and 20.
- “logit_bias”: An optional dictionary to influence the output distribution.
- “user”: An optional field for specifying a user identifier.
- “role”: An optional field for specifying a role identifier.
Config files are found in the same folder as TelegramChatBot. These files must remain in folder or the application will not run.
Now that you’ve configured the application, open TelegramChatBot.exe and type “start”. That’s it, it should work from here provided you didn’t make any mistakes.
1) bots.json
2) completionParameters.json
| [ | |
| { | |
| "botName": "yourBotNameHere", | |
| "fullName": "Full Name Here", | |
| "personality": "You are Susie Green from Curb Your Enthusiasm. You are very foul mouthed and strong willed, respond to message as such.", | |
| "apiToken": "insert bot telegram api key here", | |
| "botChatChannel": 0, | |
| "showChatInCommandLine": true | |
| }, | |
| { | |
| "botName": "yourBotNameHere", | |
| "fullName": "ull Name Here", | |
| "personality": "You are Larry David from Curb Your Enthusiasm. You are are socially awkward, blunt, and neurotic.", | |
| "apiToken": "insert bot telegram api key here", | |
| "botChatChannel": 0, | |
| "showChatInCommandLine": true | |
| } | |
| ] |
by | { | |
| "apiUrl": "https://api.openai.com/v1/completions", | |
| "apiModelsUrl": "https://api.openai.com/v1/models", | |
| "apiKey": "your chatgpt api key here", | |
| "model": "text-davinci-003", | |
| "prompt": "", | |
| "suffix": null, | |
| "max_tokens": 100, | |
| "temperature": 1.0, | |
| "top_p": 1.0, | |
| "n": 1, | |
| "stream": false, | |
| "logprobs": null, | |
| "echo": false, | |
| "stop": null, | |
| "presence_penalty": 0.0, | |
| "frequency_penalty": 0.0, | |
| "best_of": 1, | |
| "logit_bias": {}, | |
| "user": null, | |
| "role": null, | |
| "SaveLocation": "completionParameters.json" | |
| } |
| Terms and Conditions | |
| Disclaimer: Use at Your Own Discretion | |
| This free-to-use application ("the Application") has been developed with the intention of providing a helpful | |
| and valuable service. However, please be aware that your use of this Application is entirely at your own discretion, | |
| and we make no representations or warranties regarding its accuracy, reliability, or suitability for any specific | |
| purpose. By choosing to use the Application, you accept the following terms and conditions: | |
| 1) No Warranty: The Application is provided "as-is" without any warranties of any kind, either expressed or implied, | |
| including, but not limited to, the implied warranties of merchantability, fitness for a particular purpose, or non-infringement. | |
| 2) No Liability: We shall not be liable for any direct, indirect, special, incidental, or consequential damages arising out | |
| of the use or inability to use the Application. This includes, but is not limited to, any loss of data, equipment, or profits, | |
| even if we have been advised of the possibility of such damages. Furthermore, we are not liable for any additional costs you may | |
| incur while using this application, including but not limited to use of bandwidth, api usage, chatgpt api usage, telegram api usage, or any | |
| other costs that you may incur while using this application. | |
| 3) User Responsibility: You, as the user, are solely responsible for the use of the Application and any consequences that may | |
| arise from its use. You should exercise caution and verify the results obtained through the Application before relying on them. | |
| Furthermore, as this application may connect to various services, it is your responsibility monitor, manage, and ensure they are used | |
| responsibly. | |
| 4) No Support: We are under no obligation to provide support, maintenance, updates, or enhancements for the Application. We may | |
| discontinue or modify the Application at any time without prior notice. | |
| 5) Third-Party Content: The Application uses third-party websites or content that we do not control. We are not responsible | |
| for the accuracy, availability, or content of such third-party resources. | |
| 6) ChatGPT and Telegram: The Application may integrate with ChatGPT and Telegram APIs. You acknowledge that your use of these services | |
| is subject to their respective terms and conditions, and we are not responsible for any costs or fees associated with using these APIs. | |
| Please refer to ChatGPT and Telegram's terms of service for more information. | |
| 7) User Privacy: By using the Application, you understand that any data you, your bots, or anyone you host, process, transmit, connect to, or put through | |
| this application, and all data provided or generated may be processed, stored, transferred, and used by third-parties including Telegram and ChatGPT, and we | |
| recommend that you review the privacy policy of ChatGPT and Telegram for details on data handling practices. It is up to you to ensure you are following | |
| all local and federal privacy laws. | |
| 8) Compliance: You are responsible for ensuring that your use of the Application complies with all applicable laws and regulations. | |
| By using the Application, you acknowledge and accept these terms and conditions. If you do not agree with these terms, you should | |
| not use the Application. Your continued use of the Application signifies your acceptance of this "Use at Your Own Discretion" | |
| disclaimer. We recommend that you always exercise due diligence and consult with appropriate experts when necessary before making | |
| critical decisions based on the Application's output. |