big-AGI

big-AGI is a browser-based AI workspace for researchers, developers and people who want to compare model answers before relying on them. Its Beam feature sends the same prompt to several models in parallel, without letting them see each other's replies. You can examine disagreements, then use Merge to combine the responses into a single answer with a preset or custom prompt.

The open-source foundation uses the MIT license and supports self-hosting with Docker, deployment on Vercel and local development. A hosted service is also available. The app stores data on your device, but prompts sent to cloud models go to those providers. It connects to local LLM backends including Ollama, LocalAI and LM Studio, alongside Anthropic, OpenAI, Google Gemini, DeepSeek and OpenRouter. Cloud connections use your own API keys.

Personas hold reusable instructions, documents, images and memory, along with a chosen model and its settings. You can adjust reasoning effort, temperature and other controls per request or persona. The AI Inspector exposes outgoing requests, parameters, token usage and costs so you can check what the app sends.

The interface supports split-pane chats, conversation branches and message editing for exploring alternative answers. Voice input and calls sit alongside file uploads, PDF extraction, camera and screen input. It also supports image generation and renders PlantUML and Mermaid diagrams. You can use the hosted app in a browser or install it as a web app.