Exact artifacts

Validated Ollama variants

Every entry is pinned to a quantization, blob size and digest.

Q4_K_M · 2.32 GB

Phi-4 mini

phi4-mini:3.8b-q4_K_M

Compare NVIDIA GPUs → Q4_K_M · 2.33 GB

Qwen3

qwen3:4b-instruct-2507-q4_K_M

Compare NVIDIA GPUs → Q4_K_M · 4.36 GB

Qwen2.5 Coder

qwen2.5-coder:7b-instruct-q4_K_M

Compare NVIDIA GPUs → Q4_K_M · 6.96 GB

Mistral Nemo

mistral-nemo:12b-instruct-2407-q4_K_M

Compare NVIDIA GPUs → MXFP4 · 12.8 GB

GPT-OSS

gpt-oss:20b

Compare NVIDIA GPUs → Q4_K_M · 3.16 GB

Qwen3.5 4B

qwen3.5:4b-q4_K_M

Compare NVIDIA GPUs → Q4_K_M · 6.14 GB

Qwen3.5 9B

qwen3.5:9b-q4_K_M

Compare NVIDIA GPUs → Q4_K_M · 16.2 GB

Qwen3.5 27B

qwen3.5:27b-q4_K_M

Compare NVIDIA GPUs → Q4_K_M · 22.2 GB

Qwen3.5 35B-A3B

qwen3.5:35b-a3b-q4_K_M

Compare NVIDIA GPUs → Q4_K_M · 4.87 GB

DeepSeek-R1 8B

deepseek-r1:8b

Compare NVIDIA GPUs → Q4_K_M · 8.37 GB

DeepSeek-R1 14B

deepseek-r1:14b

Compare NVIDIA GPUs → Q4_K_M · 18.5 GB

DeepSeek-R1 32B

deepseek-r1:32b

Compare NVIDIA GPUs → Q4_K_M · 3.11 GB

Gemma 3 4B

gemma3:4b

Compare NVIDIA GPUs → Q4_K_M · 7.59 GB

Gemma 3 12B

gemma3:12b

Compare NVIDIA GPUs → Q4_K_M · 16.2 GB

Gemma 3 27B

gemma3:27b

Compare NVIDIA GPUs →

Canonical catalogue

Search all 700 models

Filter by name, publisher, task, size or compatibility coverage.