Deploy your private Gemini application for free with one click, supporting Gemini 1.5 Pro, Gemini 1.5 Flash, Gemini Pro and Gemini Pro Vision models.
English · 简体中文
Web App / Desktop App / Issues
Share GeminiNextChat Repository
Simple interface, supports image recognition and voice conversation
Supports Gemini 1.5 and Gemini 1.5 Flash multimodal models
Support plugins, with built-in Web search, Web reader, Arxiv search, Weather and other practical plugins
A cross-platform application client that supports a permanent menu bar, doubling your work efficiency
Note: If you encounter problems during the use of the project, you can check the known problems and solutions of FAQ.
- Deploy for free with one-click on Vercel in under 1 minute
- Provides a very small (~4MB) cross-platform client (Windows/MacOS/Linux), can stay in the menu bar to improve office efficiency
- Supports multi-modal models and can understand images, videos, audios and some text documents
- Talk mode: Let you talk directly to Gemini
- Visual recognition allows Gemini to understand the content of the picture
- Assistant market with hundreds of selected system instruction
- Support plugins, with built-in Web search, Web reader, Arxiv search, Weather and other practical plugins
- Conversation list, so you can keep track of important conversations or discuss different topics with Gemini
- Full Markdown support: LaTex formulas, code highlighting, and more
- Automatically compress contextual chat records to save Tokens while supporting very long conversations
- Privacy and security, all data is saved locally in the user's browser
- Support PWA, can run as an application
- Well-designed UI, responsive design, supports dark mode
- Extremely fast first screen loading speed, supporting streaming response
- Static deployment supports deployment on any website service that supports static pages, such as Github Page, Cloudflare, Vercel, etc.
- Multi-language support: English、简体中文、繁体中文、日本語、한국어、Español、Deutsch、Français、Português、Русский and العربية
- Reconstruct the topic square and introduce Prompt list
- Use tauri to package desktop applications
- Implementation based on functionCall plug-in
- Support conversation list
- Get Gemini API Key
- Click
- Start using
If you want to update instantly, you can check out the GitHub documentation to learn how to synchronize a forked project with upstream code.
You can star or watch this project or follow author to get release notifications in time.
Your Gemini api key. If you need to enable the server api, this is required.
Default:
https://generativelanguage.googleapis.com
Examples:
http://your-gemini-proxy.com
Override Gemini api request base url. **To avoid server-side proxy url leaks, links in front-end pages will not be overwritten. **
Default:
https://generativelanguage.googleapis.com
Example:
http://your-gemini-upload-proxy.com
Override Gemini file upload api base url. **To avoid server-side proxy url leaks, links in front-end pages will not be overwritten. **
Custom model list, default: all.
Default:
https://chat-agents.lobehub.com
Examples:
http://your-assistant-market-proxy.com
Override assistant market api request base url. The api link in the front-end interface will be adjusted synchronously.
File upload size limit. There is no file size limit by default.
Access password.
Injected script code can be used for statistics or error tracking.
Only used to set the page base path in static deployment mode.
This project provides limited access control. Please add an environment variable named ACCESS_PASSWORD on the vercel environment variables page.
After adding or modifying this environment variable, please redeploy the project for the changes to take effect.
This project supports custom model lists. Please add an environment variable named NEXT_PUBLIC_GEMINI_MODEL_LIST in the .env file or environment variables page.
The default model list is represented by all, and multiple models are separated by ,.
If you need to add a new model, please directly write the model name all,new-model-name, or use the + symbol plus the model name to add, that is, all,+new-model-name.
If you want to remove a model from the model list, use the - symbol followed by the model name to indicate removal, i.e. all,-existing-model-name. If you want to remove the default model list, you can use -all.
If you want to set a default model, you can use the @ symbol plus the model name to indicate the default model, that is, all,@default-model-name.
If you have not installed pnpm
npm install -g pnpm# 1. install nodejs and yarn first
# 2. config local variables, please change `.env.example` to `.env` or `.env.local`
# 3. run
pnpm install
pnpm devNodeJS >= 18, Docker >= 20
The Docker version needs to be 20 or above, otherwise it will prompt that the image cannot be found.
⚠️ Note: Most of the time, the docker version will lag behind the latest version by 1 to 2 days, so the "update exists" prompt will continue to appear after deployment, which is normal.
docker pull xiangfa/talk-with-gemini:latest
docker run -d --name talk-with-gemini -p 5481:3000 xiangfa/talk-with-geminiYou can also specify additional environment variables:
docker run -d --name talk-with-gemini \
-p 5481:3000 \
-e GEMINI_API_KEY=AIzaSy... \
-e ACCESS_PASSWORD=your-password \
xiangfa/talk-with-geminiIf you need to specify other environment variables, please add -e key=value to the above command to specify it.
Deploy using docker-compose.yml:
version: '3.9'
services:
talk-with-gemini:
image: xiangfa/talk-with-gemini
container_name: talk-with-gemini
environment:
- GEMINI_API_KEY=AIzaSy...
- ACCESS_PASSWORD=your-password
ports:
- 5481:3000You can also build a static page version directly, and then upload all files in the out directory to any website service that supports static pages, such as Github Page, Cloudflare, Vercel, etc..
pnpm build:exportIf you deploy the project in a subdirectory and encounter resource loading failures when accessing, please add EXPORT_BASE_PATH=/path/project in the .env file or variable setting page.
-
Use Cloudflare AI Gateway to forward APIs. Currently, Cloudflare AI Gateway already supports Google Vertex AI related APIs. For how to use it, please refer to How to Use Cloudflare AI Gateway. This solution is fast and stable, and is recommended.
-
Use Cloudflare Worker for API proxy forwarding. For detailed settings, please refer to How to Use Cloudflare Worker Proxy API. Note that this solution may not work properly in some cases.
Currently, the two models Gemini 1.5 Pro and Gemini 1.5 Flash support most images, audios, videos and some text files. For details. For other document types, we will try to use LangChain.js later.