Nehanth/pooled
Peer-to-peer LLM inference in the browser: pool your devices to run big open models, for chat and coding agents.
View on GitHub →Where this sits in the catalog
Most-wanted open issues
Ranked by 👍 reactions. 97 open issues total, checked 2026-09-29.
| Issue | 👍 | Comments | Opened |
|---|---|---|---|
TURN relay for strict NATs enhancement, P1, area: room, area: infra, impact: reliability, area: networking | 0 | 1 | 3 wk |
Prefill speed: land the prefill/* kernels (GEMM tiles, chunked DeltaNet, flash-attn prefill, MoE grouping) enhancement, P0, area: engine, area: kernels, impact: performance | 0 | 1 | 3 wk |
Automatic re-deal when a device leaves or joins enhancement, P1, area: room, area: protocol, impact: reliability | 0 | 2 | 3 wk |
Pooled CLI: npx pooled serve with an OpenAI-compatible endpoint (later) enhancement, P2, area: engine, area: infra, impact: ux | 0 | 1 | 3 wk |
Offline rooms: hotspot, QR signaling, peer weight sharing enhancement, P2, area: room, area: infra, impact: reliability, area: networking | 0 | 1 | 3 wk |
Alternatives in Browser agents
The web data API to search, scrape, and interact at scale. 🔥
Playwright MCP server undetected by anti-bots and captchas: AI agent browses the web on anti-detect stealth Fi…
Playwright MCP server undetected by anti-bots and captchas: AI agent browses the web on anti-detect stealth Fi…
Playwright MCP server undetected by anti-bots and captchas: AI agent browses the web on anti-detect stealth Fi…
Make Any Website into CLI & Use your logged-in browser by AI agent.