arXiv:2611.17753v1 Announce Type: cross Abstract: Running large language models (LLMs) on local devices keeps data private, removes per-token cost, cuts network latency, and keeps working without a connection. Two shifts now strain how such models are served. On one device, several agents share a locally served model with an interactive user. Across devices, a developer or small team often owns several capable machines from different vendors, which sit idle most of the day because every local server serves one runtime on one machine. We present Superfluid, an open-source LLM server built for local inference at both scales. On one device, Superfluid drives llama.cpp, MLX, or BaseRT in isolated worker processes through one scheduler that moves interactive requests ahead of agent work, computes a prompt shared by many requests once, and writes every token to a durable log before it reaches the client. Across devices, one Superfluid server turns a fleet of machines into a single API endpoint. Machines with different hardware serve copies of the same model, receive work according to their measured speed, and can join or fail under load without losing a request. For Qwen3.8-27B on an Apple M5 Pro, a chat sent while eight agents are working receives its first token in 1.5 s, against 291-612 s on llama-server, Ollama, and mlx_lm.server. On six machines from three chip vendors serving 64 simulated users for one hour, Superfluid answers all 5,041 requests while machines join one at a time and one is killed and rejoins; throughput grows from 112 to 534 tok/s and the median first-token time falls from 99 to 12 s. On twelve Mac Studio M3 Ultras it sustains 10,656 tok/s across 38,563 requests from 384 agents without a failure. Superfluid is available under the Apache-2.0 licence at https://github.com/basecompute/superfluid.
Superfluid: Local AI Server for Distributed LLM Inference
Basecompute released Superfluid, an open-source LLM server under the Apache-2.0 licence that pools multiple local machines into a single API endpoint, per the arXiv paper 2611.17753v1. On an Apple M5 Pro running Qwen3.8-27B, a chat sent while eight agents were working received its first token in 1.5 s, versus 291-612 s on llama-server, Ollama and mlx_lm.server. Across six machines from three chip vendors serving 64 simulated users for one hour, Superfluid answered all 5,041 requests while machines joined and one was killed and rejoined, raising throughput from 112 to 534 tok/s and cutting median first-token time from 99 to 12 s.
Run your AI side-project on zahid.host
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.