Slopweb

A local browser shell where AI generates every page as live HTML.

Quick Start

npm install -g slopweb
slopweb

Pick a local model or Codex in the launcher, then open the printed local URL.

Local models and custom endpoints

Slopweb detects common local runtimes and OpenAI-compatible APIs, including Ollama, LM Studio, llama.cpp/llamafile, vLLM, SGLang, Jan, text-generation-webui, and KoboldCpp.

slopweb models
slopweb --base-url http://localhost:11434/v1 --model llama3.2

Custom provider definitions can live in ~/.slopweb/models.json.

Server options

slopweb --port 9000
slopweb --strict-port
slopweb --lan
slopweb --no-picker

The HTTP API is localhost-only by default. Use --lan only when you intentionally want LAN access.

Run from source

pnpm install
pnpm start
pnpm run check

License

MIT