Can I run this LLM?

Which LLMs can your computer run?

All 64 tracked open models, strongest first — with the memory each needs at its smallest quantization and how many of our tracked machines fit it. Open a model to check your exact computer, or start from your machine instead.

• Kimi K3 (MoE)Kimi97≥ 862.2 GB0/69 machines
• GLM-5.3 (MoE)GLM95≥ 255.1 GB4/69 machines
• GLM-5.3 Flash (MoE)GLM93≥ 126.6 GB8/69 machines
• Kimi K2.7 (MoE)Kimi92≥ 340.5 GB3/69 machines
• GLM-5.2 (MoE)GLM91≥ 255.1 GB4/69 machines
• Qwen3.8 Flash Next (MoE)Qwen90≥ 81.8 GB14/69 machines
• Hunyuan Hy3 (MoE)Hunyuan88≥ 108.8 GB8/69 machines
• Qwen3.8 27BQwen85≥ 11.9 GB55/69 machines
• Gemma 4 31BGemma84≥ 13.7 GB53/69 machines
• DeepSeek V4 Flash (MoE)DeepSeek84≥ 98.0 GB8/69 machines
• Laguna S 2.1 (MoE)Laguna82≥ 40.8 GB27/69 machines
• Ternary Bonsai 2 27BBonsai82≥ 7.0 GB66/69 machines
• Mistral Small 4 119B (MoE)Mistral81≥ 41.1 GB27/69 machines
• Ling 3.0 Flash (MoE)Ling81≥ 46.7 GB18/69 machines
• GLM-4.5 Air (MoE)GLM80≥ 46.9 GB18/69 machines
• Ornith 1.5 35B-A3B (MoE)Ornith80≥ 14.0 GB53/69 machines
• Step 3.7 Flash (MoE)Step80≥ 72.3 GB14/69 machines
• Gemma 4 26B-A4B (MoE)Gemma79≥ 11.6 GB55/69 machines
• GPT-OSS 120B (MoE)GPT-OSS79≥ 63.5 GB18/69 machines
• Ornith 1.0 35B (MoE)Ornith78≥ 13.1 GB53/69 machines
• Qwen3.6 35B-A3B (MoE)Qwen77≥ 13.2 GB53/69 machines
• Muse Glimmer 30BMuse77≥ 12.0 GB55/69 machines
• Qwen3 Coder Next 80B-A3B (MoE)Qwen76≥ 30.1 GB29/69 machines
• Qwen3.6 27BQwen75≥ 12.9 GB53/69 machines
• GLM-4.7 Flash (MoE)GLM74≥ 12.4 GB55/69 machines
• Laguna XS 2.1 (MoE)Laguna74≥ 13.2 GB53/69 machines
• Llama 3.3 70B InstructLlama73≥ 28.5 GB30/69 machines
• North Mini Code 1.0 (MoE)North73≥ 11.7 GB55/69 machines
• Devstral Small 2 24BMistral72≥ 10.4 GB65/69 machines
• Kimi Linear 48B-A3B (MoE)Kimi72≥ 18.4 GB41/69 machines
• DeepSeek-R1-Distill 32BDeepSeek72≥ 14.2 GB53/69 machines
• Nemotron 3.5 Lightning 30B-A3B (MoE)Nemotron72≥ 19.7 GB41/69 machines
• Qwen3 Coder 30B-A3B (MoE)Qwen71≥ 12.5 GB55/69 machines
• Llama 4 Scout (MoE)Llama71≥ 41.2 GB27/69 machines
• Qwen3 30B-A3B (MoE)Qwen70≥ 12.5 GB55/69 machines
• Qwen3 32BQwen70≥ 14.2 GB53/69 machines
• Gemma 3 27BGemma70≥ 12.1 GB55/69 machines
• Granite 4.2 30BGranite69≥ 12.7 GB54/69 machines
• Gemma 4 12BGemma68≥ 5.9 GB66/69 machines
• Mistral Small 3.2 24BMistral67≥ 10.4 GB65/69 machines
• Ornith 1.5 9BOrnith67≥ 5.0 GB69/69 machines
• GPT-OSS 20B (MoE)GPT-OSS66≥ 12.4 GB55/69 machines
• DeepSeek-R1-Distill 14BDeepSeek66≥ 7.4 GB66/69 machines
• 1-bit Bonsai 27BBonsai66≥ 4.9 GB69/69 machines
• Ornith 1.0 9BOrnith65≥ 6.6 GB66/69 machines
• Qwen3 14BQwen64≥ 7.2 GB66/69 machines
• Phi-4 14BPhi64≥ 7.2 GB66/69 machines
• MiMo V2.6 Distill Qwen 9BMiMo64≥ 4.6 GB69/69 machines
• Qwen3.5 9BQwen63≥ 5.1 GB69/69 machines
• Gemma 3 12BGemma62≥ 6.2 GB66/69 machines
• DeepSeek-R1-Distill 8BDeepSeek60≥ 4.5 GB69/69 machines
• Granite 4.2 8BGranite60≥ 4.9 GB69/69 machines
• Gemma 4 E4BGemma58≥ 4.6 GB69/69 machines
• GLM-4 9BGLM58≥ 5.0 GB69/69 machines
• Qwen3 8BQwen58≥ 4.7 GB69/69 machines
• Qwen3.5 4BQwen56≥ 2.9 GB69/69 machines
• Llama 3.1 8BLlama54≥ 4.5 GB69/69 machines
• MiniCPM5 2BMiniCPM54≥ 2.0 GB69/69 machines
• LFM2.5 2.6BLFM46≥ 2.5 GB69/69 machines
• LFM2.5 8B-A1B (MoE)LFM45≥ 6.0 GB66/69 machines
• Llama 3.2 3BLlama40≥ 3.3 GB69/69 machines
• Gemma 2 2BGemma36≥ 2.9 GB69/69 machines
• LFM2.5 1.2BLFM31≥ 1.6 GB69/69 machines
• Llama 3.2 1BLlama26≥ 1.7 GB69/69 machines

capability 0–100 · memory = smallest-quant footprint at 4K context · machine counts across the tracked device catalog