Laguna-S-2.1 118B reportedly runs at 39W and 20–40 tokens per second
AlpinDale · x · 2026-07-26
- A post claims Laguna-S-2.1 (118B) is running at only 39W while producing 20–40 tok/s.
- The image shows a terminal-style readout alongside the model name and power draw.
- The message is essentially a brag about efficient local inference: large model, low wattage, usable speed.
More from Infra
- DeepSpeed shows AutoEP/AutoTP/AutoSP parallelism gains, hosts office hours — PyTorch · 2026-10-27
- Cloudflare Durable Objects validated as top architecture for AI agent WebSocket servers — tristanbob · 2026-10-27
- Gary Marcus: With Q3 data added, Anthropic's compute-efficiency growth shows deceleration — GaryMarcus · 2026-10-27
- Inference is just Kubernetes with a multi-trillion-dollar market, and the communities overlap — matt_slotnick · 2026-10-27
- vLLM eyes Rust migration as inference engines race to rewrite the stack — AccBalanced · 2026-10-27
- Resin lets an agent build standalone LLM inference engines; vLLM and SGLang optimize DeepSeek kernels differently — AccBalanced · 2026-10-27