LobeChat, now rebranded upstream as LobeHub, is an open-source workspace for talking to many AI models through one polished interface and organizing them into agent teams that work around the clock. It started as a private ChatGPT-style front end and has grown into scheduled agents, shared workspaces, and a marketplace of skills and plugins.[1]
Quick verdict
Individuals and small teams who want one self-hosted interface for OpenAI, Anthropic, Google, Qwen, and local Ollama models benefit most from LobeChat. Unified multi-provider chat with an assistant marketplace, knowledge base, and one-click deployment on operator infrastructure is the concrete capability behind that recommendation. The material constraint is operational and licensing shape: self-hosting needs API keys plus hosting, and the repository now sits under a custom LobeHub Community License whose terms must be checked. If you can carry provider API spend and confirm the license fits, LobeChat is a strong default; if you need turnkey AI with no setup, choose ChatGPT instead.[1][2][3]
Key features
One interface for hosted and local models Chat with OpenAI, Anthropic, Google, Qwen, and others next to Ollama and LM Studio endpoints, switching providers without changing interface.[1][2]
Assistant marketplace and visual agent builder Start from prebuilt assistants or assemble task-specific agents with prompts, tools, and knowledge attached.[1]
Agent groups with scheduled operation Organize agents into teams that take shifts, report on work, and run on schedules instead of single-turn chats.[1][2]
Knowledge base with file upload and RAG Upload documents, manage them centrally, and ground answers in retrieved content with vector storage on paid Cloud tiers.[1][2]
Plugin system and MCP marketplace Extend agents through function-calling plugins plus installable MCP servers, with thousands of listed skills, plugins, and tools.[1]
Multi-modal chat surface Send images with vision input, speak and listen with speech-to-text and text-to-speech, generate images, and render artifacts with visible chain-of-thought.[1]
Mobile PWA and messaging gateways Install the interface as a phone app or reach agents through Telegram, Discord, WeChat, WhatsApp, and QQ gateways.[1]
One-click self-hosting or managed Cloud Deploy on Vercel, Zeabur, Sealos, or Alibaba Cloud, run the Docker image with provider API keys, or skip infrastructure with LobeHub Cloud.[1][2]
Who will love it
Self-hosters wanting a private ChatGPT-style UI keep chats on operator infrastructure with a commercial-grade interface through bring-your-own API keys plus Docker or one-click deploys.[1]
Multi-model individuals and teams cover hosted flagships and local models in one workspace instead of a separate app and subscription per provider.[1][2]
Agent builders turn one-off chats into repeatable team workflows with groups, schedules, shared memory, and the MCP marketplace.[1][2]
Who should avoid it
Non-technical users wanting turnkey AI face API keys, model selection, and deployment or plan setup that assume comfort with configuration; they should use finished ChatGPT instead.[1]
The repository names a custom LobeHub Community License dated 2026 with unverified OSI status, so counsel must clear it before adoption.[1]
Cloud credits burn per agent interaction and self-hosted use still bills provider API usage plus hosting, so budget-capped teams must meter spend.[1]
Alternatives and compatibility
LobeChat is the natural catalogue comparison to ChatGPT: the repository positions self-deployment against the incumbent while adding multi-provider choice and self-hosting the incumbent does not offer.[1][2]
LobeChat fits readers who can carry API and hosting spend with cleared licensing and want provider breadth, agent teams, and deployment choice in one workspace.[1][2]
Readers who need managed AI with no keys, deploys, or license review should choose the incumbent instead; nobody should adopt either self-host path without owning configuration or spend.[1]
Questions worth asking
What is the best LobeChat alternative for non-technical users?
Is LobeChat really no-code?
Mostly for daily use, not for setup. Chatting, building assistants, and installing plugins are point-and-click, but connecting providers, supplying API keys, and deploying or subscribing still need configuration work.[1]
How much does LobeChat cost per month?
As checked October 9, 2026, LobeHub Cloud lists a free tier at $0 with 500,000 credits per month, paid tiers at $9.90, $19.90, and $39.90 per month headline for Starter, Premium, and Ultimate, custom Enterprise pricing, and 50 percent off for students. Self-hosting has no listed software price, but provider API usage and hosting still bill separately, so confirm current billing terms on the pricing page before budgeting.[1]
Can you self-host LobeChat?
Yes, with a Docker image and compose template, one-click templates for Vercel, Zeabur, Sealos, and Alibaba Cloud, and provider API keys supplied as environment variables. Enterprise self-hosting with a commercial license, brand theming, and private models is a custom tier, and the Community License governs the self-hosted edition, so check its terms first.[1][2]
Sources and editorial review
Source research. Reviewed on October 9, 2026. Written by Sanjib Sah. AI assisted; reviewed by a human editor.
We have not run or tested this product.
- lobehub/lobe-chat repository (now LobeHub)
official · Observed October 9, 2026
Repository homepage: project rebranded from LobeChat to LobeHub, described as Chief Agent Operator for hiring, scheduling, and reporting on an AI agent team; 7x24 agent teams. Stars ~83,081, forks 15,961, created 2023-05-21, not archived, default branch canary, 14,512 commits, principal maintainers arvinxx and canisminor1990.
- LobeHub README: features, deployment, plugins, license
official · Observed October 9, 2026
README feature and deployment sections: agent builder, agent groups, pages, schedule, project, workspace, personal memory, IM gateway (Telegram, Discord, WeChat, WhatsApp, QQ); 10,000+ skills, plugins, MCP, artifacts, tools, image generation, thinking; unified multi-model access (OpenAI, Anthropic, Google, Qwen, Ollama, LM Studio, Azure OpenAI, AWS Bedrock); self-host via Vercel one-click, Zeabur, Sealos, Alibaba Cloud, Docker image and compose with provider API-key env vars; plugin system via function calling; ecosystem (lobe-ui, icons, tts, mcp-marketplace). License section names the LobeHub Community License, copyright 2026 LobeHub.
- LobeHub pricing
official · Observed October 9, 2026
Official pricing: Free $0 (500k Cloud credits/month, 15MB file storage, 100 vector entries); paid Cloud tiers Starter, Premium, Ultimate at $9.9/$19.9/$39.9 per month headline with monthly/annual billing variance; Enterprise custom with commercial license, brand theming, user management, self-hosted provider, private models; 50% student discount; credits consumed per agent interaction.
- LobeHub repository identity and topics
official · Observed October 9, 2026
Repository identity block: Deploy your own ChatGPT positioning; lobehub organization with 83.1k followers; repository topics chatgpt, deepseek, gemini, ollama, openai, anthropic, claude, qwen, r1.