All reviews
Model reviews.
30 hands-on reviews of local LLMs, tested and benchmarked.
Snowflake Arctic Instruct — The Enterprise MoE Reviewed
Cohere Aya 23 35B — Multilingual Powerhouse Tested
Mistral Codestral 22B — A Coding LLM Review
Cohere Command R 35B — A RAG-First LLM Tested
Cohere Command R+ 104B — Enterprise RAG Champion?
Databricks DBRX Instruct — The Forgotten MoE Reviewed
DeepSeek Coder V2 Lite 16B — Local Coding Sweet Spot
DeepSeek R1 Distill 32B — Reasoning Test After 200 Prompts
DeepSeek R1 Distill 7B — Reasoning on 6 GB VRAM
Gemma 2 27B — Is Google's Local LLM Still Relevant in 2026?
Gemma 2 2B — The 2 GB Local LLM That Surprised Us
Gemma 2 9B — Tested on a $400 GPU
Llama 3.2 3B — The Best Local LLM Under 5 GB?
Llama 3.1 8B — A 2026 Reassessment
Llama 3.2 Vision 11B — Multimodal Tested on 50 Images
Llama 3.3 70B Instruct — Complete Hands-On Review (2026)
Mistral 7B Instruct — Why It Still Holds Up in 2026
Mistral Nemo 12B — Tested on Multilingual Tasks
Mistral Small 3.1 24B — A Real-World Review
Mixtral 8x7B — Still Worth Running in 2026?
Microsoft Phi-4 14B — The MIT-Licensed Reasoning Surprise
Phi-3.5 Mini — Microsoft's Pocket Powerhouse
Qwen 2.5 32B Instruct — Versus DeepSeek R1 32B
Qwen 2.5 Coder 32B — The Best Open-Source Coding Model in 2026?
Qwen 2.5 Coder 7B — A Free Copilot Replacement?
Qwen 3 14B — Best LLM for 12 GB VRAM Cards?
Qwen 3 235B-A22B — The MoE Flagship Reviewed
Qwen 3 32B — Hands-On Review After 30 Days
Qwen 3 8B — The 8 GB VRAM Sweet Spot Review
01.AI Yi 34B — A Multilingual LLM Tested