DeepSeek Runs at Only 8 tps on Mac Studio
BitXorBit · reddit · 2026-07-08
Someone locally ran DeepSeek V4 Flash on a Mac Studio M3 Ultra, noting that llama.cpp now supports the model and unsloth has released a GGUF version. The main takeaway is that the author's real-world test yielded only about 8 token/s. They shared the llama-server launch parameters, asking if this speed is normal or if there is a configuration issue.
More from Infra
- vLLM eyes Rust migration as inference engines race to rewrite the stack — AccBalanced · 2026-10-27
- Anthropic's S-1 reveals up to $567B compute spend and self-built data centers — jeremyakahn · 2026-10-27
- Resin lets an agent build standalone LLM inference engines; vLLM and SGLang optimize DeepSeek kernels differently — AccBalanced · 2026-10-27
- Local benchmark: Flint matches Qwen3.6 quality at 50/52 while running 3.6x faster on M4 Pro — RichardSocher · 2026-10-27
- Anthropic S-1: CSP marketplace revenue jumped from 11% to 48% in three years — AccBalanced · 2026-10-27
- Alchemy adds effect/Workflow support across Durable Objects, Workflows and Lambda Durable Functions — samgoodwin89 · 2026-10-27