clevis@tinyweights:/home$ cat ./welcome.txt
> Small language models like Gemma, Phi, SmolLM, and Qwen, run locally. Benchmarks, quantization, and hands-on deployment guides for small LLMs on real hardware.
> New releases tested the week they drop. Straight talk on what is worth running and what is not.
46 posts · last updated 2026-07-19 · all writing CC BY 4.0
clevis@tinyweights:/home$ ls -lh --sort=time
07-19
10min
comparisons
Jamba Reasoning 3B vs SmolLM3-3B: Which Open 3B Model to Run Locally?
07-18
10min
comparisons
Granite 4.1 8B vs LFM2.5-8B-A1B: Dense vs MoE at 8B for Local Inference
07-16
9min
comparisons
Granite 4.1 8B vs Ministral 3 8B: Which 8B Model Should You Run Locally?
07-16
9min
news
SmolLM3-3B: The Fully Open Small Language Model That Punches Way Above Its Weight
07-15
8min
comparisons
LFM2.5-1.2B-Thinking vs Qwen3.5-0.8B: Which Tiny Model Deserves Your Gigabyte?
07-14
8min
guides
Run Bonsai 27B Locally: A 27B Model in 3.9 GB, Even on a Phone
07-14
7min
comparisons
SmolLM3-3B vs Phi-4-mini: Which Small Model Should You Run Locally?
07-13
7min
comparisons
Jamba Reasoning 3B vs Phi-4-mini-reasoning: Which Small Reasoning Model Should You Run Locally?
07-11
8min
comparisons
LFM2.5-8B-A1B vs ZAYA1-8B: Which 8B MoE Reasoning Model Should You Run Locally?
07-10
8min
comparisons
North Mini Code vs Qwen3.6-35B-A3B: Which Local Coding MoE Wins?