1,174 Macs form Darkbloom inference network, growing 4.5x in 60 days at half the cost

gajesh · x · 2026-10-15

A Darkbloom provider shared real numbers from running home Macs as inference nodes for open-weight models:

• The provider plugged in a 2021 M1 Max MacBook Pro and a Mac mini M5 Pro, together serving 35B-class open models and completing 8,147 jobs for strangers on the internet
• The Darkbloom network now spans 1,174 Macs, growing 4.5x in 60 days on OpenRouter
• Gemma 4 26B is the most used model, with gpt-oss-20b close behind — both running on Macs people already own, at up to half what the biggest host charges
• The community keeps optimizing for Apple Silicon via YukonResearch challenges

A glimpse of the distributed idle-compute economy for open model inference.

Original post →

More from Infra

• patio11 on multi-agent dev: clearly the future, and we'll buy so much inference — IgorCarron · 2026-10-15
• Hugging Face Hub makes model tensor inspection up to 20x faster, DeepSeek-V3.2 drops from 9.2s to 0.45s — victormustar · 2026-10-15
• RTX 5060 Ti 16GB surges up to 144% in Japan, hitting $1,292 as RTX 5080 tops $2,400 — chemist_slime · 2026-10-15
• Fish Audio Drama 3 lands in vLLM-Omni: 128 concurrent streams on one H200 — vllm_project · 2026-10-15
• ByteDance in advanced talks with MediaTek for custom AI data-center chips, Bloomberg reports — mark_k · 2026-10-15
• Decision Models mining to support smaller models, even CPU-only like Laya — airesearch12 · 2026-10-15

Infra channel →