https://github.com/appatalks/chatgpt-html
OpenAI & Google Generative Language Chat
https://github.com/appatalks/chatgpt-html
chatgpt dalle gemini genrative-ai gpt-4 gpt-5 lm-studio o1 o1-preview o3 openai thinking
Last synced: 2 months ago
JSON representation
OpenAI & Google Generative Language Chat
- Host: GitHub
- URL: https://github.com/appatalks/chatgpt-html
- Owner: appatalks
- License: mit
- Created: 2023-01-24T04:03:45.000Z (almost 3 years ago)
- Default Branch: main
- Last Pushed: 2025-08-10T20:46:35.000Z (4 months ago)
- Last Synced: 2025-09-23T10:02:13.461Z (3 months ago)
- Topics: chatgpt, dalle, gemini, genrative-ai, gpt-4, gpt-5, lm-studio, o1, o1-preview, o3, openai, thinking
- Language: JavaScript
- Homepage:
- Size: 6.02 MB
- Stars: 26
- Watchers: 1
- Forks: 11
- Open Issues: 4
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# ChatGPT HTML - Using OpenAI APIs;
v.4.0
This allows you to easily interact with OpenAI and Google Generative APIs.
## Getting Started
1. Add your OpenAI API key to the "```OPENAI_API_KEY```" variable and (optionally) AWS, Google Keys in ```config.json``` for additional functionality.
2. Open ```index.html``` and have fun!
- If opening via file://, copy `config.local.example.js` to `config.local.js` and set your keys there (no fetch needed).
- Or, serve over http(s) and keep using `config.json`.
3. Optional: - Suno-Ai's Bark TTS Engine. Run ```server.py``` (GPU Enabled)
4. **Note: You may have to review/adjust the code for your specific env. ie ```CIDR ranges```, ```NGINX/webserver``` configuration,```scripting``` piece etc.**
5. Local usage without a server:
- Copy `config.local.example.js` to `config.local.js` and fill your keys.
- Open `index.html` directly (file://). No fetch call to config.json is needed.
## Features
- Keeps conversation memory
- OpenAI ```o1```, ```o1-*``` models
- OpenAI ```o3-mini``` model added (great for code - tier 4+ API)
- OpenAI ```gpt-4o``` models
- OpenAI `gpt-5-mini` (experimental) and `latest` alias (per OpenAI latest-model guidance)
- Latest Google Gemini 2.0 ```Thinking``` model
- [lmstudio API](https://lmstudio.ai/docs/api/openai-api) local models
- Dall-E Image Generation
- Google Vision
- Model Selection, Multiple languages, and Print Conversation.
- Convert to Speech using Amazon Polly's Text-to-Speech service.
- Suno-Ai's Bark TTS Engine available
- Use Google Search with the Keyword "Google"
- Images served with Google Image Search
- Additional scraped data with scripts
- Basic Error handling
## Contributing
See `.github/copilot-instructions.md` for contribution guidance and model wiring conventions.
## Bugs
- Check Issues
- Response with ```"usage":{"completion_tokens":420}``` causes weird display bug on-screen.
- **Not for Production use (really messy code, likely security concerns, all-over-the-place, good playgroud and learning tho!)**
Grabbed the inital idea from here https://www.codeproject.com/Articles/5350454/Chat-GPT-in-JavaScript
Complete overhaul of the code base.