What is LobeChat?
If you are looking for a modern, open-source AI chat client and conversation framework, LobeChat is arguably the best choice available today. Maintained by the LobeHub team under the permissive MIT license, this incredible project has already amassed over 40,000 GitHub stars. It has rapidly become a favorite among developers and AI enthusiasts worldwide.
LobeChat is much more than just another simple chat UI. It natively supports over 100 models, including industry leaders like OpenAI, Claude, Gemini, DeepSeek, and Qwen. The platform comes packed with advanced features such as TTS (Text-to-Speech) voice synthesis, STT (Speech-to-Text) voice recognition, and a highly extensible plugin market. The user interface includes seamless multi-session management, flawless Markdown rendering, and beautiful code syntax highlighting. Furthermore, it fully supports PWA (Progressive Web App) installation, mobile adaptation, and both dark and light themes. You can utilize it as a personal AI assistant, a frontend for your team's internal knowledge base, or a unified entry point for all your diverse AI models.
Why Self-Host LobeChat?
As an indie developer and cloud infrastructure enthusiast, I often get asked: "Why not just use the official hosted version or a third-party SaaS?" The primary answer is absolute control. Self-hosting LobeChat gives you complete ownership of your data, prompts, and conversation history.
For startup founders and DevOps engineers, it provides a unified, customizable interface for your team without worrying about external API rate limits or data privacy compliance issues. For AI hobbyists, it is the perfect sandbox to test different models, tweak parameters, and experiment with new plugins. Because the architecture is incredibly lightweight—consisting of a Next.js frontend and backend API routes—it is remarkably easy and cheap to deploy on your own cloud infrastructure.
Server Requirements & Config Guide
LobeChat is designed to be highly efficient. The pure frontend combined with a lightweight backend proxy means you do not need expensive, high-end hardware to get started.
Here is a quick comparison of server configurations to help you choose:
| Configuration | CPU / RAM | Storage | Best For |
|---|---|---|---|
| Minimum | 1 Core / 1 GB | 20 GB | Personal use, static + light proxy |
| Recommended | 2 Cores / 2 GB | 40 GB | Multi-user, plugins, TTS/STT features |
| Production | 4 Cores / 8 GB+ | 100 GB+ | High traffic, enterprise team knowledge base |
Cloud Server Pricing Hooks (September 2026)
When choosing a cloud provider, lightweight application servers or ECS instances are your best bet. Here are the current deals I highly recommend:
Alibaba Cloud
Alibaba Cloud offers a lightweight 2 Core 2GB server with 40G ESSD and 200M peak bandwidth for just ¥38 (approx. $5.30) per year (prices subject to official promotion pages) during their flash sales at 10:00 and 15:00 daily. If you need a traditional ECS, the economy 'e' instance (2 Core 2GB, 3M bandwidth) is ¥99 per year (prices subject to official promotion pages) with the same renewal price. For absolute minimal needs, a 1 Core 1GB lightweight server is under ¥10 per month (prices subject to official promotion pages). You can check these deals at https://www.aliyun.com/minisite/goods?userCode=tzlh4rrj
Tencent Cloud
Tencent Cloud is running a massive procurement season event. New users can grab a lightweight server starting at ¥38 (approx. $5.30) per year (prices subject to official promotion pages), with a "buy 1 year get 3 months" offer and same-price renewals. Their 2 Core 2GB with 4M bandwidth is also ¥99 per year (prices subject to official promotion pages). Note that this event ends on October 12, 2026. Grab them here: https://cloud.tencent.com/act/cps/redirect?redirect=1003&cps_key=U4Wwh5F3y1uS8pJz
One-Click Cloud Deployment Options
Both major cloud providers offer one-click deployment solutions to make your life easier. Alibaba Cloud Compute Nest and Tencent Cloud HAI (High-Availability Instance) often feature pre-configured images for popular open-source projects.
If an official LobeChat image is available in the marketplace, you can simply select it during checkout, and the server will be provisioned with the application already running. However, if no official image is currently listed, I highly recommend the universal Docker deployment path. It is cleaner, easier to update, and gives you full control over the environment variables and configurations.
Docker / Docker Compose Universal Deployment
Deploying to a cloud server via Docker is the most robust method. It makes configuring custom domains and HTTPS much simpler. Here are the exact steps I use in my own workflows.
Step 1: Install Docker
First, SSH into your cloud server and install Docker. For Ubuntu or Debian, you can use the official convenience script:
curl -fsSL https://get.docker.com | sh
systemctl enable docker && systemctl start docker
Step 2: Run the LobeChat Container
Once Docker is installed, pull and run the LobeChat image. This command maps port 3210 and runs the container in the background:
docker run -d -p 3210:3210 --name lobechat lobehub/lobe-chat
(Note: If you prefer Docker Compose, you can create a docker-compose.yml file and use docker compose up -d with the same image and port mappings.)
Step 3: Access the Interface
Open your browser and visit http://server-ip:3210 to start using the application. You should immediately see the beautiful, responsive LobeChat welcome screen.
Step 4: Configure API Keys
Navigate to the Settings page in the left sidebar. Enter your OpenAI, Claude, or other model API keys. Because LobeChat supports 100+ models, you can easily switch between providers right from the UI without changing any code.
Step 5: Set Up Nginx Reverse Proxy (Recommended)
For production use, it is highly recommended to set up an Nginx reverse proxy with a custom domain and HTTPS. This secures your API keys in transit and provides a professional URL. You can use Certbot to generate free Let's Encrypt SSL certificates for your domain.
Common Questions & Troubleshooting
Q: I get a 502 Bad Gateway error when accessing the IP.
A: This usually means the container is still starting up or has crashed. Check the logs using docker logs lobechat. Also, ensure port 3210 is explicitly opened in your cloud provider's security group or firewall settings.
Q: Can I use my own database for persistent storage?
A: Yes, LobeChat supports a server database mode. You will need to configure environment variables in your Docker run command or docker-compose.yml to connect to an external PostgreSQL database.
Q: The TTS and STT features are not working or crashing.
A: These features require specific API keys and sufficient server resources. Ensure you are using the recommended 2 Core 2GB configuration, as audio processing and plugin executions can be memory-intensive.
Cost-Saving Tips
To get the most out of your deployment without breaking the bank, follow these practical strategies:
- Leverage New-User Discounts: Both Alibaba and Tencent Cloud offer massive discounts for first-time buyers. The ¥38 (approx. $5.30) per year deals (prices subject to official promotion pages) are unbeatable for personal projects.
- Choose the Right Tier: If you are just testing or using it for personal daily chats, the minimum 1 Core 1GB configuration is perfectly fine. Only upgrade to 2 Core 2GB if you plan to use TTS/STT or host it for a small team.
- Lock in Renewal Prices: The ECS and lightweight servers mentioned above often come with "same-price renewal" guarantees. This protects you from sudden price hikes in the second year.
Official Deal Links
Ready to deploy your own AI assistant? Here are the direct links to the current promotions:
Alibaba Cloud September Deals:
https://www.aliyun.com/minisite/goods?userCode=tzlh4rrj
Tencent Cloud Procurement Season:
https://cloud.tencent.com/act/cps/redirect?redirect=1003&cps_key=U4Wwh5F3y1uS8pJz
Please comply with model service provider terms and local laws and regulations.
Top comments (0)