← Dashboard  /  News

2026-09-09

AI/Tech Digest 2026-09-09

AI & local LLM

1. Kimi K3 (2.8T) at 1 token/s on a MacBook Pro, streamed from four SSDs [LAB]
   Demonstrates extreme local inference performance for a massive 2.8T parameter model using novel SSD streaming techniques.
   https://github.com/argonautlabsai/deltafin (github.com)
2. Qwen/Qwen-Drive-1.0-4B · Hugging Face [LAB]
   A new 4B parameter fine-tuned model from Qwen, relevant to the reader's interest in small-to-mid-sized model architectures.
   https://www.reddit.com/r/LocalLLaMA/comments/1wauxg9/qwenqwendrive104b_hugging_face/ (reddit.com)
3. Mistral raises €3B to make sovereign, open-weight AI the technology frontier [LAB]
   A major structural industry event involving a key vendor of open-weight models that shapes the local-inference landscape.
   https://mistral.ai/news/mistral-makes-sovereign-open-weight-ai-to-frontier/ (mistral.ai)