Search

… The smaller models coming out of Google, Alibaba, and a few others are actually built for the hardware most people already own , so you don't need a home server rack or a 24GB GPU to have something useful running on your machine. …

Jul 20, 2026 · Nolen Jonker

What 'distilling' an AI model actually means and why it matters to self-hosting open LLMs

… Companies such as Meta, Google, Alibaba, and DeepSeek have enough resources to continue training models without distillation. …

Aug 10, 2026 · Anurag Singh

I stopped running the biggest local LLM that could fit, and a 2B model handles 90% of what I need

… Qwen 3.5 2B came out a few weeks earlier in March 2026 from Alibaba's Qwen team, as part of the small series alongside the 0.8B, 4B, and 9B. …

Jul 2, 2026 · Nolen Jonker

I tested a local LLM against a frontier cloud model, and the gap was smaller than I expected

["qwen3","alibaba","gpt-5"]

Jul 1, 2026 · Adam Conway

Followed topics

Alibaba

I ran Gemma, Llama, and Qwen on my phone, and only one of them is worth the storage

I trimmed down my local AI stack, and I should have done it sooner

I stopped using Qwen and Gemma after finding a local LLM that actually thinks before answering

I split my coding work between Claude, Qwen3-Coder and Gemma 4, and it costs less than paying for one subscription

I ran Gemma 4 and Qwen 3.5 for the same local tasks, and one pulled miles ahead

I run a 24GB GPU instead of paying for Claude or Codex, and Qwen 3.6 keeps up more than I expected

I tested every tiny local LLM worth running, and only three survived the cut

What 'distilling' an AI model actually means and why it matters to self-hosting open LLMs

I stopped running the biggest local LLM that could fit, and a 2B model handles 90% of what I need

I tested a local LLM against a frontier cloud model, and the gap was smaller than I expected