{ "dataset": { "name": "RigRunsAI", "compiled": "2026-10-15", "license": "CC-BY-4.0", "licenseUrl": "https://creativecommons.org/licenses/by/4.0/", "defaultContext": 8192, "notes": [ "Model weight bytes are Hugging Face file sizes for a single GGUF, not estimates from bits-per-weight.", "Architecture fields come from config.json, or from the GGUF header when the official config is access-gated.", "Parameter counts are the gguf.total value on the Hugging Face model API for the cited GGUF repo." ] }, "models": [ { "slug": "llama-3.2-1b", "name": "Llama 3.2 1B Instruct", "family": "Llama", "maker": "Meta", "hfCreated": "2024-09-18", "license": "llama3.2", "parameters": 1235814432, "maxContext": 131072, "ollama": "llama3.2:1b", "role": "small", "blurb": "1B dense instruct model. The architecture numbers are the GGUF header, because Meta's config.json is access-gated.", "configUrl": "https://huggingface.co/bartowski/Llama-3.2-1B-Instruct-GGUF/blob/main/Llama-3.2-1B-Instruct-Q4_K_M.gguf", "configNote": "GGUF metadata: llama.block_count, embedding_length, head counts, key_length, context_length. Parameter count is gguf.total from the Hugging Face model API.", "ggufRepo": "bartowski/Llama-3.2-1B-Instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 807694464, "repo": "bartowski/Llama-3.2-1B-Instruct-GGUF", "file": "Llama-3.2-1B-Instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Llama-3.2-1B-Instruct-GGUF/blob/main/Llama-3.2-1B-Instruct-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 911503488, "repo": "bartowski/Llama-3.2-1B-Instruct-GGUF", "file": "Llama-3.2-1B-Instruct-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/Llama-3.2-1B-Instruct-GGUF/blob/main/Llama-3.2-1B-Instruct-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 1021800576, "repo": "bartowski/Llama-3.2-1B-Instruct-GGUF", "file": "Llama-3.2-1B-Instruct-Q6_K.gguf", "url": "https://huggingface.co/bartowski/Llama-3.2-1B-Instruct-GGUF/blob/main/Llama-3.2-1B-Instruct-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 1321083008, "repo": "bartowski/Llama-3.2-1B-Instruct-GGUF", "file": "Llama-3.2-1B-Instruct-Q8_0.gguf", "url": "https://huggingface.co/bartowski/Llama-3.2-1B-Instruct-GGUF/blob/main/Llama-3.2-1B-Instruct-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 16, "fullLayers": 16, "hidden": 2048, "heads": 32, "kvHeads": 8, "headDim": 64, "intermediate": 8192, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false, "vocab": null }, "multimodal": false }, { "slug": "llama-3.2-3b", "name": "Llama 3.2 3B Instruct", "family": "Llama", "maker": "Meta", "hfCreated": "2024-09-18", "license": "llama3.2", "parameters": 3212749888, "maxContext": 131072, "ollama": "llama3.2:3b", "role": "small", "blurb": "3B dense instruct model. Architecture from the public GGUF header.", "configUrl": "https://huggingface.co/bartowski/Llama-3.2-3B-Instruct-GGUF/blob/main/Llama-3.2-3B-Instruct-Q4_K_M.gguf", "configNote": "GGUF metadata on the Q4_K_M file. Meta's config.json is access-gated.", "ggufRepo": "bartowski/Llama-3.2-3B-Instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 2019377696, "repo": "bartowski/Llama-3.2-3B-Instruct-GGUF", "file": "Llama-3.2-3B-Instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Llama-3.2-3B-Instruct-GGUF/blob/main/Llama-3.2-3B-Instruct-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 2322154016, "repo": "bartowski/Llama-3.2-3B-Instruct-GGUF", "file": "Llama-3.2-3B-Instruct-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/Llama-3.2-3B-Instruct-GGUF/blob/main/Llama-3.2-3B-Instruct-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 2643853856, "repo": "bartowski/Llama-3.2-3B-Instruct-GGUF", "file": "Llama-3.2-3B-Instruct-Q6_K.gguf", "url": "https://huggingface.co/bartowski/Llama-3.2-3B-Instruct-GGUF/blob/main/Llama-3.2-3B-Instruct-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 3421899296, "repo": "bartowski/Llama-3.2-3B-Instruct-GGUF", "file": "Llama-3.2-3B-Instruct-Q8_0.gguf", "url": "https://huggingface.co/bartowski/Llama-3.2-3B-Instruct-GGUF/blob/main/Llama-3.2-3B-Instruct-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 28, "fullLayers": 28, "hidden": 3072, "heads": 24, "kvHeads": 8, "headDim": 128, "intermediate": 8192, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false, "vocab": null }, "multimodal": false }, { "slug": "llama-3.1-8b", "name": "Llama 3.1 8B Instruct", "family": "Llama", "maker": "Meta", "hfCreated": "2024-07-18", "license": "llama3.1", "parameters": 8030261312, "maxContext": 131072, "ollama": "llama3.1:8b", "role": "general", "blurb": "8B dense instruct model and the usual baseline for a gaming PC.", "configUrl": "https://huggingface.co/bartowski/Meta-Llama-3.1-8B-Instruct-GGUF/blob/main/Meta-Llama-3.1-8B-Instruct-Q4_K_M.gguf", "configNote": "GGUF metadata. head_dim is embedding_length / head_count (4096 / 32). Meta's config.json is access-gated.", "ggufRepo": "bartowski/Meta-Llama-3.1-8B-Instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 4920739232, "repo": "bartowski/Meta-Llama-3.1-8B-Instruct-GGUF", "file": "Meta-Llama-3.1-8B-Instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Meta-Llama-3.1-8B-Instruct-GGUF/blob/main/Meta-Llama-3.1-8B-Instruct-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 5732992416, "repo": "bartowski/Meta-Llama-3.1-8B-Instruct-GGUF", "file": "Meta-Llama-3.1-8B-Instruct-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/Meta-Llama-3.1-8B-Instruct-GGUF/blob/main/Meta-Llama-3.1-8B-Instruct-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 6596011424, "repo": "bartowski/Meta-Llama-3.1-8B-Instruct-GGUF", "file": "Meta-Llama-3.1-8B-Instruct-Q6_K.gguf", "url": "https://huggingface.co/bartowski/Meta-Llama-3.1-8B-Instruct-GGUF/blob/main/Meta-Llama-3.1-8B-Instruct-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 8540775840, "repo": "bartowski/Meta-Llama-3.1-8B-Instruct-GGUF", "file": "Meta-Llama-3.1-8B-Instruct-Q8_0.gguf", "url": "https://huggingface.co/bartowski/Meta-Llama-3.1-8B-Instruct-GGUF/blob/main/Meta-Llama-3.1-8B-Instruct-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 32, "fullLayers": 32, "hidden": 4096, "heads": 32, "kvHeads": 8, "headDim": 128, "intermediate": 14336, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false, "vocab": null }, "multimodal": false }, { "slug": "llama-3.3-70b", "name": "Llama 3.3 70B Instruct", "family": "Llama", "maker": "Meta", "hfCreated": "2024-11-26", "license": "llama3.3", "parameters": 70553706560, "maxContext": 131072, "ollama": "llama3.3:70b", "role": "general", "blurb": "70B dense instruct model. Only the single-file Q4_K_M size is listed; larger quants were split archives.", "configUrl": "https://huggingface.co/bartowski/Llama-3.3-70B-Instruct-GGUF/blob/main/Llama-3.3-70B-Instruct-Q4_K_M.gguf", "configNote": "GGUF metadata. Meta's config.json is access-gated.", "ggufRepo": "bartowski/Llama-3.3-70B-Instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 42520398816, "repo": "bartowski/Llama-3.3-70B-Instruct-GGUF", "file": "Llama-3.3-70B-Instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Llama-3.3-70B-Instruct-GGUF/blob/main/Llama-3.3-70B-Instruct-Q4_K_M.gguf", "recommended": true } ], "arch": { "kind": "dense", "layers": 80, "fullLayers": 80, "hidden": 8192, "heads": 64, "kvHeads": 8, "headDim": 128, "intermediate": 28672, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false, "vocab": null }, "multimodal": false }, { "slug": "qwen2.5-7b", "name": "Qwen2.5 7B Instruct", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2024-09-16", "license": "apache-2.0", "parameters": 7615616512, "maxContext": 32768, "ollama": "qwen2.5:7b", "blurb": "Dense Qwen2.5 instruct model. use_sliding_window is false in config.json, so every layer keeps a full KV cache.", "configUrl": "https://huggingface.co/Qwen/Qwen2.5-7B-Instruct/raw/main/config.json", "configNote": "Official config.json. head_dim = hidden_size / num_attention_heads.", "ggufRepo": "bartowski/Qwen2.5-7B-Instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 4683074240, "repo": "bartowski/Qwen2.5-7B-Instruct-GGUF", "file": "Qwen2.5-7B-Instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-7B-Instruct-GGUF/blob/main/Qwen2.5-7B-Instruct-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 5444831936, "repo": "bartowski/Qwen2.5-7B-Instruct-GGUF", "file": "Qwen2.5-7B-Instruct-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-7B-Instruct-GGUF/blob/main/Qwen2.5-7B-Instruct-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 6254199488, "repo": "bartowski/Qwen2.5-7B-Instruct-GGUF", "file": "Qwen2.5-7B-Instruct-Q6_K.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-7B-Instruct-GGUF/blob/main/Qwen2.5-7B-Instruct-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 8098525888, "repo": "bartowski/Qwen2.5-7B-Instruct-GGUF", "file": "Qwen2.5-7B-Instruct-Q8_0.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-7B-Instruct-GGUF/blob/main/Qwen2.5-7B-Instruct-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 28, "fullLayers": 28, "hidden": 3584, "heads": 28, "kvHeads": 4, "headDim": 128, "intermediate": 18944, "vocab": 152064, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false, "role": "general" }, { "slug": "qwen2.5-14b", "name": "Qwen2.5 14B Instruct", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2024-09-16", "license": "apache-2.0", "parameters": 14770033664, "maxContext": 32768, "ollama": "qwen2.5:14b", "blurb": "Dense Qwen2.5 instruct model. use_sliding_window is false in config.json, so every layer keeps a full KV cache.", "configUrl": "https://huggingface.co/Qwen/Qwen2.5-14B-Instruct/raw/main/config.json", "configNote": "Official config.json. head_dim = hidden_size / num_attention_heads.", "ggufRepo": "bartowski/Qwen2.5-14B-Instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 8988110976, "repo": "bartowski/Qwen2.5-14B-Instruct-GGUF", "file": "Qwen2.5-14B-Instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-14B-Instruct-GGUF/blob/main/Qwen2.5-14B-Instruct-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 10508873856, "repo": "bartowski/Qwen2.5-14B-Instruct-GGUF", "file": "Qwen2.5-14B-Instruct-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-14B-Instruct-GGUF/blob/main/Qwen2.5-14B-Instruct-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 12124684416, "repo": "bartowski/Qwen2.5-14B-Instruct-GGUF", "file": "Qwen2.5-14B-Instruct-Q6_K.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-14B-Instruct-GGUF/blob/main/Qwen2.5-14B-Instruct-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 15701598336, "repo": "bartowski/Qwen2.5-14B-Instruct-GGUF", "file": "Qwen2.5-14B-Instruct-Q8_0.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-14B-Instruct-GGUF/blob/main/Qwen2.5-14B-Instruct-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 48, "fullLayers": 48, "hidden": 5120, "heads": 40, "kvHeads": 8, "headDim": 128, "intermediate": 13824, "vocab": 152064, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false, "role": "general" }, { "slug": "qwen2.5-32b", "name": "Qwen2.5 32B Instruct", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2024-09-17", "license": "apache-2.0", "parameters": 32763876352, "maxContext": 32768, "ollama": "qwen2.5:32b", "blurb": "Dense Qwen2.5 instruct model. use_sliding_window is false in config.json, so every layer keeps a full KV cache.", "configUrl": "https://huggingface.co/Qwen/Qwen2.5-32B-Instruct/raw/main/config.json", "configNote": "Official config.json. head_dim = hidden_size / num_attention_heads.", "ggufRepo": "bartowski/Qwen2.5-32B-Instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 19851336576, "repo": "bartowski/Qwen2.5-32B-Instruct-GGUF", "file": "Qwen2.5-32B-Instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-32B-Instruct-GGUF/blob/main/Qwen2.5-32B-Instruct-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 23262157696, "repo": "bartowski/Qwen2.5-32B-Instruct-GGUF", "file": "Qwen2.5-32B-Instruct-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-32B-Instruct-GGUF/blob/main/Qwen2.5-32B-Instruct-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 26886155136, "repo": "bartowski/Qwen2.5-32B-Instruct-GGUF", "file": "Qwen2.5-32B-Instruct-Q6_K.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-32B-Instruct-GGUF/blob/main/Qwen2.5-32B-Instruct-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 34820885376, "repo": "bartowski/Qwen2.5-32B-Instruct-GGUF", "file": "Qwen2.5-32B-Instruct-Q8_0.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-32B-Instruct-GGUF/blob/main/Qwen2.5-32B-Instruct-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 64, "fullLayers": 64, "hidden": 5120, "heads": 40, "kvHeads": 8, "headDim": 128, "intermediate": 27648, "vocab": 152064, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false, "role": "general" }, { "slug": "qwen2.5-72b", "name": "Qwen2.5 72B Instruct", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2024-09-16", "license": "other", "parameters": 72706203648, "maxContext": 32768, "ollama": "qwen2.5:72b", "blurb": "Dense 72B instruct model. The Hugging Face license field is other. Only the single-file Q4_K_M size is listed.", "configUrl": "https://huggingface.co/Qwen/Qwen2.5-72B-Instruct/raw/main/config.json", "configNote": "Official config.json. head_dim = hidden_size / num_attention_heads.", "ggufRepo": "bartowski/Qwen2.5-72B-Instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 47415715488, "repo": "bartowski/Qwen2.5-72B-Instruct-GGUF", "file": "Qwen2.5-72B-Instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Qwen2.5-72B-Instruct-GGUF/blob/main/Qwen2.5-72B-Instruct-Q4_K_M.gguf", "recommended": true } ], "arch": { "kind": "dense", "layers": 80, "fullLayers": 80, "hidden": 8192, "heads": 64, "kvHeads": 8, "headDim": 128, "intermediate": 29568, "vocab": 152064, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false, "role": "general" }, { "slug": "qwen3-8b", "name": "Qwen3 8B", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2025-04-27", "license": "apache-2.0", "parameters": 8190735360, "maxContext": 40960, "ollama": "qwen3:8b", "blurb": "Dense Qwen3 model with a 40,960-token context. Sliding window is off in config.json.", "configUrl": "https://huggingface.co/Qwen/Qwen3-8B/raw/main/config.json", "configNote": "Official config.json, including head_dim.", "ggufRepo": "unsloth/Qwen3-8B-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 5027784512, "repo": "unsloth/Qwen3-8B-GGUF", "file": "Qwen3-8B-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3-8B-GGUF/blob/main/Qwen3-8B-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 5851113280, "repo": "unsloth/Qwen3-8B-GGUF", "file": "Qwen3-8B-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3-8B-GGUF/blob/main/Qwen3-8B-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 6725900096, "repo": "unsloth/Qwen3-8B-GGUF", "file": "Qwen3-8B-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Qwen3-8B-GGUF/blob/main/Qwen3-8B-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 8709519168, "repo": "unsloth/Qwen3-8B-GGUF", "file": "Qwen3-8B-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Qwen3-8B-GGUF/blob/main/Qwen3-8B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 36, "fullLayers": 36, "hidden": 4096, "heads": 32, "kvHeads": 8, "headDim": 128, "intermediate": 12288, "vocab": 151936, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false, "role": "general" }, { "slug": "qwen3-14b", "name": "Qwen3 14B", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2025-04-27", "license": "apache-2.0", "parameters": 14768307200, "maxContext": 40960, "ollama": "qwen3:14b", "blurb": "Dense Qwen3 model with a 40,960-token context. Sliding window is off in config.json.", "configUrl": "https://huggingface.co/Qwen/Qwen3-14B/raw/main/config.json", "configNote": "Official config.json, including head_dim.", "ggufRepo": "unsloth/Qwen3-14B-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 9001753984, "repo": "unsloth/Qwen3-14B-GGUF", "file": "Qwen3-14B-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3-14B-GGUF/blob/main/Qwen3-14B-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 10514570624, "repo": "unsloth/Qwen3-14B-GGUF", "file": "Qwen3-14B-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3-14B-GGUF/blob/main/Qwen3-14B-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 12121938304, "repo": "unsloth/Qwen3-14B-GGUF", "file": "Qwen3-14B-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Qwen3-14B-GGUF/blob/main/Qwen3-14B-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 15698534784, "repo": "unsloth/Qwen3-14B-GGUF", "file": "Qwen3-14B-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Qwen3-14B-GGUF/blob/main/Qwen3-14B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 40, "fullLayers": 40, "hidden": 5120, "heads": 40, "kvHeads": 8, "headDim": 128, "intermediate": 17408, "vocab": 151936, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false, "role": "general" }, { "slug": "qwen3-32b", "name": "Qwen3 32B", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2025-04-27", "license": "apache-2.0", "parameters": 32762123264, "maxContext": 40960, "ollama": "qwen3:32b", "blurb": "Dense Qwen3 model with a 40,960-token context. Sliding window is off in config.json.", "configUrl": "https://huggingface.co/Qwen/Qwen3-32B/raw/main/config.json", "configNote": "Official config.json, including head_dim.", "ggufRepo": "unsloth/Qwen3-32B-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 19762150048, "repo": "unsloth/Qwen3-32B-GGUF", "file": "Qwen3-32B-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3-32B-GGUF/blob/main/Qwen3-32B-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 23214832288, "repo": "unsloth/Qwen3-32B-GGUF", "file": "Qwen3-32B-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3-32B-GGUF/blob/main/Qwen3-32B-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 26883307168, "repo": "unsloth/Qwen3-32B-GGUF", "file": "Qwen3-32B-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Qwen3-32B-GGUF/blob/main/Qwen3-32B-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 34817719968, "repo": "unsloth/Qwen3-32B-GGUF", "file": "Qwen3-32B-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Qwen3-32B-GGUF/blob/main/Qwen3-32B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 64, "fullLayers": 64, "hidden": 5120, "heads": 64, "kvHeads": 8, "headDim": 128, "intermediate": 25600, "vocab": 151936, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false, "role": "general" }, { "slug": "qwen3-30b-a3b", "name": "Qwen3 30B-A3B", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2025-04-27", "license": "apache-2.0", "parameters": 30532122624, "maxContext": 40960, "ollama": "qwen3:30b-a3b", "role": "general", "blurb": "Mixture-of-experts Qwen3. All 48 layers are sparse (decoder_sparse_step 1) with 8 of 128 experts active.", "configUrl": "https://huggingface.co/Qwen/Qwen3-30B-A3B/raw/main/config.json", "configNote": "Official config.json: num_experts 128, num_experts_per_tok 8, moe_intermediate_size 768. No shared expert field.", "ggufRepo": "unsloth/Qwen3-30B-A3B-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 18556686912, "repo": "unsloth/Qwen3-30B-A3B-GGUF", "file": "Qwen3-30B-A3B-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3-30B-A3B-GGUF/blob/main/Qwen3-30B-A3B-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 21725581888, "repo": "unsloth/Qwen3-30B-A3B-GGUF", "file": "Qwen3-30B-A3B-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3-30B-A3B-GGUF/blob/main/Qwen3-30B-A3B-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 25092532800, "repo": "unsloth/Qwen3-30B-A3B-GGUF", "file": "Qwen3-30B-A3B-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Qwen3-30B-A3B-GGUF/blob/main/Qwen3-30B-A3B-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 32483932736, "repo": "unsloth/Qwen3-30B-A3B-GGUF", "file": "Qwen3-30B-A3B-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Qwen3-30B-A3B-GGUF/blob/main/Qwen3-30B-A3B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "moe", "layers": 48, "fullLayers": 48, "hidden": 2048, "heads": 32, "kvHeads": 4, "headDim": 128, "intermediate": 6144, "vocab": 151936, "experts": 128, "expertsPerTok": 8, "moeIntermediate": 768, "tieEmbeddings": false, "linear": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0 }, "multimodal": false }, { "slug": "qwen3.5-9b", "name": "Qwen3.5 9B", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2026-02-27", "license": "apache-2.0", "parameters": 8953803264, "maxContext": 262144, "ollama": "qwen3.5:9b", "multimodal": true, "blurb": "Hybrid model: 24 linear-attention layers and 8 full-attention layers. The GGUF used here is the language-model file.", "configUrl": "https://huggingface.co/Qwen/Qwen3.5-9B/raw/main/config.json", "configNote": "text_config.layer_types, head_dim 256, linear_* fields, mamba_ssm_dtype float32.", "ggufRepo": "unsloth/Qwen3.5-9B-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 5680522464, "repo": "unsloth/Qwen3.5-9B-GGUF", "file": "Qwen3.5-9B-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3.5-9B-GGUF/blob/main/Qwen3.5-9B-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 6577841376, "repo": "unsloth/Qwen3.5-9B-GGUF", "file": "Qwen3.5-9B-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3.5-9B-GGUF/blob/main/Qwen3.5-9B-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 7458301152, "repo": "unsloth/Qwen3.5-9B-GGUF", "file": "Qwen3.5-9B-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Qwen3.5-9B-GGUF/blob/main/Qwen3.5-9B-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 9527502048, "repo": "unsloth/Qwen3.5-9B-GGUF", "file": "Qwen3.5-9B-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Qwen3.5-9B-GGUF/blob/main/Qwen3.5-9B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "hybrid-linear", "layers": 32, "fullLayers": 8, "hidden": 4096, "heads": 16, "kvHeads": 4, "headDim": 256, "intermediate": 12288, "vocab": 248320, "linear": { "layers": 24, "keyHeads": 16, "valueHeads": 32, "keyDim": 128, "valueDim": 128, "convKernel": 4, "stateBytes": 4 }, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "role": "general" }, { "slug": "qwen3.8-27b", "name": "Qwen3.8 27B", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2026-08-05", "license": "apache-2.0", "parameters": 27320697856, "maxContext": 262144, "multimodal": true, "blurb": "Hybrid 27B model: 48 linear-attention layers and 16 full-attention layers. Recommended file is Unsloth UD-Q4_K_M.", "configUrl": "https://huggingface.co/Qwen/Qwen3.8-27B/raw/main/config.json", "configNote": "text_config layer_types and linear attention dims. No verified Ollama tag for this exact revision.", "ggufRepo": "unsloth/Qwen3.8-27B-GGUF", "quants": [ { "id": "UD-Q4_K_M", "bytes": 16464440224, "repo": "unsloth/Qwen3.8-27B-GGUF", "file": "Qwen3.8-27B-UD-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/blob/main/Qwen3.8-27B-UD-Q4_K_M.gguf", "recommended": true }, { "id": "UD-Q5_K_M", "bytes": 19771509664, "repo": "unsloth/Qwen3.8-27B-GGUF", "file": "Qwen3.8-27B-UD-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/blob/main/Qwen3.8-27B-UD-Q5_K_M.gguf", "recommended": false }, { "id": "UD-Q6_K", "bytes": 21983677344, "repo": "unsloth/Qwen3.8-27B-GGUF", "file": "Qwen3.8-27B-UD-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/blob/main/Qwen3.8-27B-UD-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 29047086048, "repo": "unsloth/Qwen3.8-27B-GGUF", "file": "Qwen3.8-27B-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Qwen3.8-27B-GGUF/blob/main/Qwen3.8-27B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "hybrid-linear", "layers": 64, "fullLayers": 16, "hidden": 5120, "heads": 24, "kvHeads": 4, "headDim": 256, "intermediate": 17408, "vocab": 248320, "linear": { "layers": 48, "keyHeads": 16, "valueHeads": 48, "keyDim": 128, "valueDim": 128, "convKernel": 4, "stateBytes": 4 }, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "ollama": null, "role": "general" }, { "slug": "qwen3.6-35b-a3b", "name": "Qwen3.6 35B-A3B", "family": "Qwen", "maker": "Alibaba", "hfCreated": "2026-04-15", "license": "apache-2.0", "parameters": 34660610688, "maxContext": 262144, "multimodal": true, "blurb": "Hybrid MoE: 256 experts, 8 active, plus a shared expert. 30 layers are linear attention and 10 are full attention.", "configUrl": "https://huggingface.co/Qwen/Qwen3.6-35B-A3B/raw/main/config.json", "configNote": "text_config num_experts 256, num_experts_per_tok 8, shared_expert_intermediate_size 512, moe_intermediate_size 512.", "ggufRepo": "unsloth/Qwen3.6-35B-A3B-GGUF", "quants": [ { "id": "UD-Q4_K_M", "bytes": 22134528992, "repo": "unsloth/Qwen3.6-35B-A3B-GGUF", "file": "Qwen3.6-35B-A3B-UD-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3.6-35B-A3B-GGUF/blob/main/Qwen3.6-35B-A3B-UD-Q4_K_M.gguf", "recommended": true }, { "id": "UD-Q5_K_M", "bytes": 26456194016, "repo": "unsloth/Qwen3.6-35B-A3B-GGUF", "file": "Qwen3.6-35B-A3B-UD-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Qwen3.6-35B-A3B-GGUF/blob/main/Qwen3.6-35B-A3B-UD-Q5_K_M.gguf", "recommended": false }, { "id": "UD-Q6_K", "bytes": 29308320736, "repo": "unsloth/Qwen3.6-35B-A3B-GGUF", "file": "Qwen3.6-35B-A3B-UD-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Qwen3.6-35B-A3B-GGUF/blob/main/Qwen3.6-35B-A3B-UD-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 36903140320, "repo": "unsloth/Qwen3.6-35B-A3B-GGUF", "file": "Qwen3.6-35B-A3B-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Qwen3.6-35B-A3B-GGUF/blob/main/Qwen3.6-35B-A3B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "hybrid-linear-moe", "layers": 40, "fullLayers": 10, "hidden": 2048, "heads": 16, "kvHeads": 2, "headDim": 256, "vocab": 248320, "experts": 256, "expertsPerTok": 8, "moeIntermediate": 512, "sharedExpertIntermediate": 512, "tieEmbeddings": false, "linear": { "layers": 30, "keyHeads": 16, "valueHeads": 32, "keyDim": 128, "valueDim": 128, "convKernel": 4, "stateBytes": 4 }, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "intermediate": null }, "ollama": null, "role": "general" }, { "slug": "gemma-3-4b", "name": "Gemma 3 4B Instruct", "family": "Gemma", "maker": "Google", "hfCreated": "2025-02-20", "license": "gemma", "parameters": 3880099328, "maxContext": 131072, "ollama": "gemma3:4b", "multimodal": true, "blurb": "Sliding-window model. Config sliding_window_pattern is 6, so every 6th layer is full attention and the rest keep a 1,024-token window.", "configUrl": "https://huggingface.co/unsloth/gemma-3-4b-it/raw/main/config.json", "configNote": "Google's config.json is gated. Fields were read from the public unsloth config and checked against the bartowski GGUF header (block_count, heads, key_length, sliding_window).", "ggufRepo": "bartowski/google_gemma-3-4b-it-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 2489758112, "repo": "bartowski/google_gemma-3-4b-it-GGUF", "file": "google_gemma-3-4b-it-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-4b-it-GGUF/blob/main/google_gemma-3-4b-it-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 2829562272, "repo": "bartowski/google_gemma-3-4b-it-GGUF", "file": "google_gemma-3-4b-it-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-4b-it-GGUF/blob/main/google_gemma-3-4b-it-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 3190604192, "repo": "bartowski/google_gemma-3-4b-it-GGUF", "file": "google_gemma-3-4b-it-Q6_K.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-4b-it-GGUF/blob/main/google_gemma-3-4b-it-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 4130226592, "repo": "bartowski/google_gemma-3-4b-it-GGUF", "file": "google_gemma-3-4b-it-Q8_0.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-4b-it-GGUF/blob/main/google_gemma-3-4b-it-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "sliding", "layers": 34, "fullLayers": 5, "slidingLayers": 29, "slidingWindow": 1024, "hidden": 2560, "heads": 8, "kvHeads": 4, "headDim": 256, "intermediate": 10240, "vocab": 262208, "tieEmbeddings": true, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null }, "role": "general" }, { "slug": "gemma-3-12b", "name": "Gemma 3 12B Instruct", "family": "Gemma", "maker": "Google", "hfCreated": "2025-03-01", "license": "gemma", "parameters": 11765788416, "maxContext": 131072, "ollama": "gemma3:12b", "multimodal": true, "blurb": "Sliding-window model. Config sliding_window_pattern is 6, so every 6th layer is full attention and the rest keep a 1,024-token window.", "configUrl": "https://huggingface.co/unsloth/gemma-3-12b-it/raw/main/config.json", "configNote": "Google's config.json is gated. Fields were read from the public unsloth config and checked against the bartowski GGUF header (block_count, heads, key_length, sliding_window).", "ggufRepo": "bartowski/google_gemma-3-12b-it-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 7300575264, "repo": "bartowski/google_gemma-3-12b-it-GGUF", "file": "google_gemma-3-12b-it-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-12b-it-GGUF/blob/main/google_gemma-3-12b-it-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 8444833824, "repo": "bartowski/google_gemma-3-12b-it-GGUF", "file": "google_gemma-3-12b-it-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-12b-it-GGUF/blob/main/google_gemma-3-12b-it-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 9660608544, "repo": "bartowski/google_gemma-3-12b-it-GGUF", "file": "google_gemma-3-12b-it-Q6_K.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-12b-it-GGUF/blob/main/google_gemma-3-12b-it-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 12509949984, "repo": "bartowski/google_gemma-3-12b-it-GGUF", "file": "google_gemma-3-12b-it-Q8_0.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-12b-it-GGUF/blob/main/google_gemma-3-12b-it-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "sliding", "layers": 48, "fullLayers": 8, "slidingLayers": 40, "slidingWindow": 1024, "hidden": 3840, "heads": 16, "kvHeads": 8, "headDim": 256, "intermediate": 15360, "vocab": 262208, "tieEmbeddings": true, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null }, "role": "general" }, { "slug": "gemma-3-27b", "name": "Gemma 3 27B Instruct", "family": "Gemma", "maker": "Google", "hfCreated": "2025-03-01", "license": "gemma", "parameters": 27009002240, "maxContext": 131072, "ollama": "gemma3:27b", "multimodal": true, "blurb": "Sliding-window model. Config sliding_window_pattern is 6, so every 6th layer is full attention and the rest keep a 1,024-token window.", "configUrl": "https://huggingface.co/unsloth/gemma-3-27b-it/raw/main/config.json", "configNote": "Google's config.json is gated. Fields were read from the public unsloth config and checked against the bartowski GGUF header (block_count, heads, key_length, sliding_window).", "ggufRepo": "bartowski/google_gemma-3-27b-it-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 16546404992, "repo": "bartowski/google_gemma-3-27b-it-GGUF", "file": "google_gemma-3-27b-it-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-27b-it-GGUF/blob/main/google_gemma-3-27b-it-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 19271391872, "repo": "bartowski/google_gemma-3-27b-it-GGUF", "file": "google_gemma-3-27b-it-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-27b-it-GGUF/blob/main/google_gemma-3-27b-it-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 22166690432, "repo": "bartowski/google_gemma-3-27b-it-GGUF", "file": "google_gemma-3-27b-it-Q6_K.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-27b-it-GGUF/blob/main/google_gemma-3-27b-it-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 28707605120, "repo": "bartowski/google_gemma-3-27b-it-GGUF", "file": "google_gemma-3-27b-it-Q8_0.gguf", "url": "https://huggingface.co/bartowski/google_gemma-3-27b-it-GGUF/blob/main/google_gemma-3-27b-it-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "sliding", "layers": 62, "fullLayers": 10, "slidingLayers": 52, "slidingWindow": 1024, "hidden": 5376, "heads": 32, "kvHeads": 16, "headDim": 128, "intermediate": 21504, "vocab": 262208, "tieEmbeddings": true, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null }, "role": "general" }, { "slug": "gemma-4-e2b", "name": "Gemma 4 E2B", "family": "Gemma", "maker": "Google", "hfCreated": "2026-03-02", "license": "apache-2.0", "parameters": 4647450147, "maxContext": 131072, "ollama": "gemma4:e2b", "multimodal": true, "blurb": "Small Gemma 4 model. 28 sliding layers (window 512) and 7 full layers. Sliding layers use head_dim and num_key_value_heads. Full layers use global_head_dim. When num_global_key_value_heads is null, full layers keep num_key_value_heads.", "configUrl": "https://huggingface.co/google/gemma-4-E2B-it/raw/main/config.json", "configNote": "Official text_config: layer_types, sliding_window, head_dim, global_head_dim, num_key_value_heads, num_global_key_value_heads.", "ggufRepo": "unsloth/gemma-4-E2B-it-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 3106738272, "repo": "unsloth/gemma-4-E2B-it-GGUF", "file": "gemma-4-E2B-it-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF/blob/main/gemma-4-E2B-it-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 3356037216, "repo": "unsloth/gemma-4-E2B-it-GGUF", "file": "gemma-4-E2B-it-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF/blob/main/gemma-4-E2B-it-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 4501721184, "repo": "unsloth/gemma-4-E2B-it-GGUF", "file": "gemma-4-E2B-it-Q6_K.gguf", "url": "https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF/blob/main/gemma-4-E2B-it-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 5048352864, "repo": "unsloth/gemma-4-E2B-it-GGUF", "file": "gemma-4-E2B-it-Q8_0.gguf", "url": "https://huggingface.co/unsloth/gemma-4-E2B-it-GGUF/blob/main/gemma-4-E2B-it-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "sliding", "layers": 35, "fullLayers": 7, "slidingLayers": 28, "slidingWindow": 512, "hidden": 1536, "heads": 8, "kvHeads": 1, "headDim": 256, "globalHeadDim": 512, "globalKvHeads": null, "intermediate": 6144, "vocab": 262144, "tieEmbeddings": true, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null }, "role": "general" }, { "slug": "gemma-4-e4b", "name": "Gemma 4 E4B", "family": "Gemma", "maker": "Google", "hfCreated": "2026-03-02", "license": "apache-2.0", "parameters": 7518069290, "maxContext": 131072, "ollama": "gemma4:e4b", "multimodal": true, "blurb": "Gemma 4 model sized for laptops and handhelds. 35 sliding layers (window 512) and 7 full layers. Sliding layers use head_dim and num_key_value_heads. Full layers use global_head_dim. When num_global_key_value_heads is null, full layers keep num_key_value_heads.", "configUrl": "https://huggingface.co/google/gemma-4-E4B-it/raw/main/config.json", "configNote": "Official text_config: layer_types, sliding_window, head_dim, global_head_dim, num_key_value_heads, num_global_key_value_heads.", "ggufRepo": "unsloth/gemma-4-E4B-it-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 4977171584, "repo": "unsloth/gemma-4-E4B-it-GGUF", "file": "gemma-4-E4B-it-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-E4B-it-GGUF/blob/main/gemma-4-E4B-it-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 5481798784, "repo": "unsloth/gemma-4-E4B-it-GGUF", "file": "gemma-4-E4B-it-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-E4B-it-GGUF/blob/main/gemma-4-E4B-it-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 7074929792, "repo": "unsloth/gemma-4-E4B-it-GGUF", "file": "gemma-4-E4B-it-Q6_K.gguf", "url": "https://huggingface.co/unsloth/gemma-4-E4B-it-GGUF/blob/main/gemma-4-E4B-it-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 8192953472, "repo": "unsloth/gemma-4-E4B-it-GGUF", "file": "gemma-4-E4B-it-Q8_0.gguf", "url": "https://huggingface.co/unsloth/gemma-4-E4B-it-GGUF/blob/main/gemma-4-E4B-it-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "sliding", "layers": 42, "fullLayers": 7, "slidingLayers": 35, "slidingWindow": 512, "hidden": 2560, "heads": 8, "kvHeads": 2, "headDim": 256, "globalHeadDim": 512, "globalKvHeads": null, "intermediate": 10240, "vocab": 262144, "tieEmbeddings": true, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null }, "role": "general" }, { "slug": "gemma-4-12b", "name": "Gemma 4 12B", "family": "Gemma", "maker": "Google", "hfCreated": "2026-05-23", "license": "apache-2.0", "parameters": 11907350576, "maxContext": 262144, "ollama": "gemma4:12b", "multimodal": true, "blurb": "Dense Gemma 4 12B. 40 sliding layers (window 1,024) and 8 full layers with global_head_dim 512 and 1 global KV head. Sliding layers use head_dim and num_key_value_heads. Full layers use global_head_dim. When num_global_key_value_heads is null, full layers keep num_key_value_heads.", "configUrl": "https://huggingface.co/google/gemma-4-12B-it/raw/main/config.json", "configNote": "Official text_config: layer_types, sliding_window, head_dim, global_head_dim, num_key_value_heads, num_global_key_value_heads.", "ggufRepo": "unsloth/gemma-4-12b-it-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 7121861440, "repo": "unsloth/gemma-4-12b-it-GGUF", "file": "gemma-4-12b-it-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-12b-it-GGUF/blob/main/gemma-4-12b-it-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 8413576000, "repo": "unsloth/gemma-4-12b-it-GGUF", "file": "gemma-4-12b-it-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-12b-it-GGUF/blob/main/gemma-4-12b-it-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 9786022720, "repo": "unsloth/gemma-4-12b-it-GGUF", "file": "gemma-4-12b-it-Q6_K.gguf", "url": "https://huggingface.co/unsloth/gemma-4-12b-it-GGUF/blob/main/gemma-4-12b-it-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 12669647680, "repo": "unsloth/gemma-4-12b-it-GGUF", "file": "gemma-4-12b-it-Q8_0.gguf", "url": "https://huggingface.co/unsloth/gemma-4-12b-it-GGUF/blob/main/gemma-4-12b-it-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "sliding", "layers": 48, "fullLayers": 8, "slidingLayers": 40, "slidingWindow": 1024, "hidden": 3840, "heads": 16, "kvHeads": 8, "headDim": 256, "globalHeadDim": 512, "globalKvHeads": 1, "intermediate": 15360, "vocab": 262144, "tieEmbeddings": true, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null }, "role": "general" }, { "slug": "gemma-4-26b-a4b", "name": "Gemma 4 26B-A4B", "family": "Gemma", "maker": "Google", "hfCreated": "2026-03-11", "license": "apache-2.0", "parameters": 25233142046, "maxContext": 262144, "ollama": "gemma4:26b-a4b", "multimodal": true, "blurb": "MoE Gemma 4. config top_k_experts is 8 of num_experts 128, moe_intermediate_size 704. 25 sliding layers and 5 full layers.", "configUrl": "https://huggingface.co/google/gemma-4-26B-A4B-it/raw/main/config.json", "configNote": "Official text_config: layer_types, sliding_window, head_dim, global_head_dim, num_key_value_heads, num_global_key_value_heads.", "ggufRepo": "unsloth/gemma-4-26b-a4b-it-GGUF", "quants": [ { "id": "UD-Q4_K_M", "bytes": 16947541728, "repo": "unsloth/gemma-4-26b-a4b-it-GGUF", "file": "gemma-4-26B-A4B-it-UD-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-26b-a4b-it-GGUF/blob/main/gemma-4-26B-A4B-it-UD-Q4_K_M.gguf", "recommended": true }, { "id": "UD-Q5_K_M", "bytes": 21150365408, "repo": "unsloth/gemma-4-26b-a4b-it-GGUF", "file": "gemma-4-26B-A4B-it-UD-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-26b-a4b-it-GGUF/blob/main/gemma-4-26B-A4B-it-UD-Q5_K_M.gguf", "recommended": false }, { "id": "UD-Q6_K", "bytes": 23172478688, "repo": "unsloth/gemma-4-26b-a4b-it-GGUF", "file": "gemma-4-26B-A4B-it-UD-Q6_K.gguf", "url": "https://huggingface.co/unsloth/gemma-4-26b-a4b-it-GGUF/blob/main/gemma-4-26B-A4B-it-UD-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 26859861728, "repo": "unsloth/gemma-4-26b-a4b-it-GGUF", "file": "gemma-4-26B-A4B-it-Q8_0.gguf", "url": "https://huggingface.co/unsloth/gemma-4-26b-a4b-it-GGUF/blob/main/gemma-4-26B-A4B-it-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "sliding-moe", "layers": 30, "fullLayers": 5, "slidingLayers": 25, "slidingWindow": 1024, "hidden": 2816, "heads": 16, "kvHeads": 8, "headDim": 256, "globalHeadDim": 512, "globalKvHeads": 2, "vocab": 262144, "tieEmbeddings": true, "experts": 128, "expertsPerTok": 8, "moeIntermediate": 704, "linear": null, "sharedExpertIntermediate": null, "intermediate": null }, "role": "general" }, { "slug": "gemma-4-31b", "name": "Gemma 4 31B", "family": "Gemma", "maker": "Google", "hfCreated": "2026-03-11", "license": "apache-2.0", "parameters": 30697345596, "maxContext": 262144, "ollama": "gemma4:31b", "multimodal": true, "blurb": "Dense Gemma 4 31B. 50 sliding layers (window 1,024) and 10 full layers.", "configUrl": "https://huggingface.co/google/gemma-4-31B-it/raw/main/config.json", "configNote": "Official text_config: layer_types, sliding_window, head_dim, global_head_dim, num_key_value_heads, num_global_key_value_heads.", "ggufRepo": "unsloth/gemma-4-31b-it-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 18323733440, "repo": "unsloth/gemma-4-31b-it-GGUF", "file": "gemma-4-31B-it-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-31b-it-GGUF/blob/main/gemma-4-31B-it-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 21658401728, "repo": "unsloth/gemma-4-31b-it-GGUF", "file": "gemma-4-31B-it-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/gemma-4-31b-it-GGUF/blob/main/gemma-4-31B-it-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 25201486784, "repo": "unsloth/gemma-4-31b-it-GGUF", "file": "gemma-4-31B-it-Q6_K.gguf", "url": "https://huggingface.co/unsloth/gemma-4-31b-it-GGUF/blob/main/gemma-4-31B-it-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 32635677632, "repo": "unsloth/gemma-4-31b-it-GGUF", "file": "gemma-4-31B-it-Q8_0.gguf", "url": "https://huggingface.co/unsloth/gemma-4-31b-it-GGUF/blob/main/gemma-4-31B-it-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "sliding", "layers": 60, "fullLayers": 10, "slidingLayers": 50, "slidingWindow": 1024, "hidden": 5376, "heads": 32, "kvHeads": 16, "headDim": 256, "globalHeadDim": 512, "globalKvHeads": 4, "intermediate": 21504, "vocab": 262144, "tieEmbeddings": true, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null }, "role": "general" }, { "slug": "gpt-oss-20b", "name": "gpt-oss-20b", "family": "gpt-oss", "maker": "OpenAI", "hfCreated": "2025-08-04", "license": "apache-2.0", "parameters": 20914757184, "maxContext": 131072, "ollama": "gpt-oss:20b", "role": "reasoning", "blurb": "Open-weight reasoning MoE. The published GGUF is MXFP4, not Q4_K_M. 4 of 32 experts run each token. Half the layers use a 128-token sliding window.", "configUrl": "https://huggingface.co/openai/gpt-oss-20b/raw/main/config.json", "configNote": "num_local_experts 32, num_experts_per_tok 4, intermediate_size 2880, sliding_window 128, layer_types.", "ggufRepo": "ggml-org/gpt-oss-20b-GGUF", "quants": [ { "id": "MXFP4", "bytes": 12109566624, "repo": "ggml-org/gpt-oss-20b-GGUF", "file": "gpt-oss-20b-MXFP4.gguf", "url": "https://huggingface.co/ggml-org/gpt-oss-20b-GGUF/blob/main/gpt-oss-20b-MXFP4.gguf", "recommended": true } ], "arch": { "kind": "sliding-moe", "layers": 24, "fullLayers": 12, "slidingLayers": 12, "slidingWindow": 128, "hidden": 2880, "heads": 64, "kvHeads": 8, "headDim": 64, "intermediate": 2880, "vocab": 201088, "experts": 32, "expertsPerTok": 4, "moeIntermediate": 2880, "tieEmbeddings": false, "linear": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null }, "multimodal": false }, { "slug": "gpt-oss-120b", "name": "gpt-oss-120b", "family": "gpt-oss", "maker": "OpenAI", "hfCreated": "2025-08-04", "license": "apache-2.0", "parameters": 116829156672, "maxContext": 131072, "ollama": "gpt-oss:120b", "role": "reasoning", "blurb": "Larger gpt-oss MoE. MXFP4 file. 4 of 128 experts run each token.", "configUrl": "https://huggingface.co/openai/gpt-oss-120b/raw/main/config.json", "configNote": "num_local_experts 128, num_experts_per_tok 4, sliding_window 128.", "ggufRepo": "ggml-org/gpt-oss-120b-GGUF", "quants": [ { "id": "MXFP4", "bytes": 63387346208, "repo": "ggml-org/gpt-oss-120b-GGUF", "file": "gpt-oss-120b-MXFP4.gguf", "url": "https://huggingface.co/ggml-org/gpt-oss-120b-GGUF/blob/main/gpt-oss-120b-MXFP4.gguf", "recommended": true } ], "arch": { "kind": "sliding-moe", "layers": 36, "fullLayers": 18, "slidingLayers": 18, "slidingWindow": 128, "hidden": 2880, "heads": 64, "kvHeads": 8, "headDim": 64, "intermediate": 2880, "vocab": 201088, "experts": 128, "expertsPerTok": 4, "moeIntermediate": 2880, "tieEmbeddings": false, "linear": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null }, "multimodal": false }, { "slug": "phi-4", "name": "Phi-4", "family": "Phi", "maker": "Microsoft", "hfCreated": "2024-12-11", "license": "mit", "parameters": 14659507200, "maxContext": 16384, "ollama": "phi4:14b", "role": "reasoning", "blurb": "Dense 14B-class Phi-4. head_dim is hidden_size / num_attention_heads from config.json (5120 / 40).", "configUrl": "https://huggingface.co/microsoft/phi-4/raw/main/config.json", "configNote": "Official config.json. Context 16,384.", "ggufRepo": "bartowski/phi-4-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 9053114816, "repo": "bartowski/phi-4-GGUF", "file": "phi-4-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/phi-4-GGUF/blob/main/phi-4-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 10604188096, "repo": "bartowski/phi-4-GGUF", "file": "phi-4-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/phi-4-GGUF/blob/main/phi-4-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 12030251456, "repo": "bartowski/phi-4-GGUF", "file": "phi-4-Q6_K.gguf", "url": "https://huggingface.co/bartowski/phi-4-GGUF/blob/main/phi-4-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 15580500416, "repo": "bartowski/phi-4-GGUF", "file": "phi-4-Q8_0.gguf", "url": "https://huggingface.co/bartowski/phi-4-GGUF/blob/main/phi-4-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 40, "fullLayers": 40, "hidden": 5120, "heads": 40, "kvHeads": 10, "headDim": 128, "intermediate": 17920, "vocab": 100352, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false }, { "slug": "phi-4-mini", "name": "Phi-4-mini Instruct", "family": "Phi", "maker": "Microsoft", "hfCreated": "2025-02-19", "license": "mit", "parameters": 3836021856, "maxContext": 131072, "ollama": "phi4-mini", "role": "small", "blurb": "Small Phi-4. The config sliding_window is 262,144, larger than the 131,072 context, so the KV estimate treats every layer as full attention.", "configUrl": "https://huggingface.co/microsoft/Phi-4-mini-instruct/raw/main/config.json", "configNote": "Official config.json. head_dim = 3072 / 24.", "ggufRepo": "bartowski/microsoft_Phi-4-mini-instruct-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 2491874688, "repo": "bartowski/microsoft_Phi-4-mini-instruct-GGUF", "file": "microsoft_Phi-4-mini-instruct-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF/blob/main/microsoft_Phi-4-mini-instruct-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 2848128384, "repo": "bartowski/microsoft_Phi-4-mini-instruct-GGUF", "file": "microsoft_Phi-4-mini-instruct-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF/blob/main/microsoft_Phi-4-mini-instruct-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 3155623296, "repo": "bartowski/microsoft_Phi-4-mini-instruct-GGUF", "file": "microsoft_Phi-4-mini-instruct-Q6_K.gguf", "url": "https://huggingface.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF/blob/main/microsoft_Phi-4-mini-instruct-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 4084611456, "repo": "bartowski/microsoft_Phi-4-mini-instruct-GGUF", "file": "microsoft_Phi-4-mini-instruct-Q8_0.gguf", "url": "https://huggingface.co/bartowski/microsoft_Phi-4-mini-instruct-GGUF/blob/main/microsoft_Phi-4-mini-instruct-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 32, "fullLayers": 32, "hidden": 3072, "heads": 24, "kvHeads": 8, "headDim": 128, "intermediate": 8192, "vocab": 200064, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false }, { "slug": "mistral-nemo-12b", "name": "Mistral Nemo 12B Instruct", "family": "Mistral", "maker": "Mistral", "hfCreated": "2024-07-17", "license": "apache-2.0", "parameters": 12247782400, "maxContext": 131072, "ollama": "mistral-nemo:12b", "blurb": "Dense 12B Mistral. Context used here is config max_position_embeddings, 131,072.", "configUrl": "https://huggingface.co/mistralai/Mistral-Nemo-Instruct-2407/raw/main/config.json", "configNote": "Official config.json head_dim 128. The GGUF header context field is larger; this site uses the config value.", "ggufRepo": "bartowski/Mistral-Nemo-Instruct-2407-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 7477208192, "repo": "bartowski/Mistral-Nemo-Instruct-2407-GGUF", "file": "Mistral-Nemo-Instruct-2407-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/Mistral-Nemo-Instruct-2407-GGUF/blob/main/Mistral-Nemo-Instruct-2407-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 8727635072, "repo": "bartowski/Mistral-Nemo-Instruct-2407-GGUF", "file": "Mistral-Nemo-Instruct-2407-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/Mistral-Nemo-Instruct-2407-GGUF/blob/main/Mistral-Nemo-Instruct-2407-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 10056213632, "repo": "bartowski/Mistral-Nemo-Instruct-2407-GGUF", "file": "Mistral-Nemo-Instruct-2407-Q6_K.gguf", "url": "https://huggingface.co/bartowski/Mistral-Nemo-Instruct-2407-GGUF/blob/main/Mistral-Nemo-Instruct-2407-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 13022372992, "repo": "bartowski/Mistral-Nemo-Instruct-2407-GGUF", "file": "Mistral-Nemo-Instruct-2407-Q8_0.gguf", "url": "https://huggingface.co/bartowski/Mistral-Nemo-Instruct-2407-GGUF/blob/main/Mistral-Nemo-Instruct-2407-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 40, "fullLayers": 40, "hidden": 5120, "heads": 32, "kvHeads": 8, "headDim": 128, "intermediate": 14336, "vocab": 131072, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false, "role": "general" }, { "slug": "mistral-small-3.2-24b", "name": "Mistral Small 3.2 24B Instruct", "family": "Mistral", "maker": "Mistral", "hfCreated": "2025-06-19", "license": "apache-2.0", "parameters": 23572403200, "maxContext": 131072, "multimodal": true, "blurb": "Dense 24B Mistral Small 3.2 (2506). Ollama's mistral-small:24b tag is labeled 2501, so this page does not show an Ollama command.", "configUrl": "https://huggingface.co/mistralai/Mistral-Small-3.2-24B-Instruct-2506/raw/main/config.json", "configNote": "text/llm config hidden 5120, 40 layers, 8 KV heads, head_dim 128.", "ggufRepo": "unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 14333922848, "repo": "unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF", "file": "Mistral-Small-3.2-24B-Instruct-2506-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/blob/main/Mistral-Small-3.2-24B-Instruct-2506-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 16763997728, "repo": "unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF", "file": "Mistral-Small-3.2-24B-Instruct-2506-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/blob/main/Mistral-Small-3.2-24B-Instruct-2506-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 19345952288, "repo": "unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF", "file": "Mistral-Small-3.2-24B-Instruct-2506-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/blob/main/Mistral-Small-3.2-24B-Instruct-2506-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 25054793248, "repo": "unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF", "file": "Mistral-Small-3.2-24B-Instruct-2506-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF/blob/main/Mistral-Small-3.2-24B-Instruct-2506-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 40, "fullLayers": 40, "hidden": 5120, "heads": 32, "kvHeads": 8, "headDim": 128, "intermediate": 32768, "vocab": 131072, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "ollama": null, "role": "general" }, { "slug": "devstral-small-2-24b", "name": "Devstral Small 2 24B Instruct", "family": "Mistral", "maker": "Mistral", "hfCreated": "2025-11-28", "license": "apache-2.0", "parameters": 23572403200, "maxContext": 393216, "role": "coding", "multimodal": true, "blurb": "Mistral's 24B coding model (2512), 40 dense layers, context 393,216. Ollama's devstral:24b tag is the earlier 2505 build, so no command is shown.", "configUrl": "https://huggingface.co/mistralai/Devstral-Small-2-24B-Instruct-2512/raw/main/config.json", "configNote": "Official config max_position_embeddings 393216, head_dim 128, 40 layers.", "ggufRepo": "unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 14334446752, "repo": "unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF", "file": "Devstral-Small-2-24B-Instruct-2512-Q4_K_M.gguf", "url": "https://huggingface.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF/blob/main/Devstral-Small-2-24B-Instruct-2512-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 16764521632, "repo": "unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF", "file": "Devstral-Small-2-24B-Instruct-2512-Q5_K_M.gguf", "url": "https://huggingface.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF/blob/main/Devstral-Small-2-24B-Instruct-2512-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 19346476192, "repo": "unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF", "file": "Devstral-Small-2-24B-Instruct-2512-Q6_K.gguf", "url": "https://huggingface.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF/blob/main/Devstral-Small-2-24B-Instruct-2512-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 25055317152, "repo": "unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF", "file": "Devstral-Small-2-24B-Instruct-2512-Q8_0.gguf", "url": "https://huggingface.co/unsloth/Devstral-Small-2-24B-Instruct-2512-GGUF/blob/main/Devstral-Small-2-24B-Instruct-2512-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 40, "fullLayers": 40, "hidden": 5120, "heads": 32, "kvHeads": 8, "headDim": 128, "intermediate": 32768, "vocab": 131072, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "ollama": null }, { "slug": "deepseek-r1-distill-14b", "name": "DeepSeek-R1 Distill Qwen 14B", "family": "DeepSeek", "maker": "DeepSeek", "hfCreated": "2025-01-20", "license": "mit", "parameters": 14770033664, "maxContext": 131072, "ollama": "deepseek-r1:14b", "role": "reasoning", "blurb": "Reasoning distill on the Qwen2.5 14B architecture. use_sliding_window is false.", "configUrl": "https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-14B/raw/main/config.json", "configNote": "Official config.json, Qwen2 architecture.", "ggufRepo": "bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 8988110240, "repo": "bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF", "file": "DeepSeek-R1-Distill-Qwen-14B-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF/blob/main/DeepSeek-R1-Distill-Qwen-14B-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 10508873120, "repo": "bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF", "file": "DeepSeek-R1-Distill-Qwen-14B-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF/blob/main/DeepSeek-R1-Distill-Qwen-14B-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 12124683680, "repo": "bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF", "file": "DeepSeek-R1-Distill-Qwen-14B-Q6_K.gguf", "url": "https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF/blob/main/DeepSeek-R1-Distill-Qwen-14B-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 15701597600, "repo": "bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF", "file": "DeepSeek-R1-Distill-Qwen-14B-Q8_0.gguf", "url": "https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-14B-GGUF/blob/main/DeepSeek-R1-Distill-Qwen-14B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 48, "fullLayers": 48, "hidden": 5120, "heads": 40, "kvHeads": 8, "headDim": 128, "intermediate": 13824, "vocab": 152064, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false }, { "slug": "deepseek-r1-distill-32b", "name": "DeepSeek-R1 Distill Qwen 32B", "family": "DeepSeek", "maker": "DeepSeek", "hfCreated": "2025-01-20", "license": "mit", "parameters": 32763876352, "maxContext": 131072, "ollama": "deepseek-r1:32b", "role": "reasoning", "blurb": "Reasoning distill on the Qwen2.5 32B architecture.", "configUrl": "https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Qwen-32B/raw/main/config.json", "configNote": "Official config.json.", "ggufRepo": "bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 19851335840, "repo": "bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF", "file": "DeepSeek-R1-Distill-Qwen-32B-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF/blob/main/DeepSeek-R1-Distill-Qwen-32B-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 23262156960, "repo": "bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF", "file": "DeepSeek-R1-Distill-Qwen-32B-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF/blob/main/DeepSeek-R1-Distill-Qwen-32B-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 26886154400, "repo": "bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF", "file": "DeepSeek-R1-Distill-Qwen-32B-Q6_K.gguf", "url": "https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF/blob/main/DeepSeek-R1-Distill-Qwen-32B-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 34820884352, "repo": "bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF", "file": "DeepSeek-R1-Distill-Qwen-32B-Q8_0.gguf", "url": "https://huggingface.co/bartowski/DeepSeek-R1-Distill-Qwen-32B-GGUF/blob/main/DeepSeek-R1-Distill-Qwen-32B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 64, "fullLayers": 64, "hidden": 5120, "heads": 40, "kvHeads": 8, "headDim": 128, "intermediate": 27648, "vocab": 152064, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "multimodal": false }, { "slug": "smollm3-3b", "name": "SmolLM3 3B", "family": "SmolLM", "maker": "Hugging Face", "hfCreated": "2025-07-08", "license": "apache-2.0", "parameters": 3075098624, "maxContext": 65536, "role": "small", "blurb": "Dense 3B model from Hugging Face. All 36 layers are full attention. No Ollama library tag was found for SmolLM3.", "configUrl": "https://huggingface.co/HuggingFaceTB/SmolLM3-3B/raw/main/config.json", "configNote": "Official config.json. head_dim = 2048 / 16.", "ggufRepo": "bartowski/HuggingFaceTB_SmolLM3-3B-GGUF", "quants": [ { "id": "Q4_K_M", "bytes": 1915305792, "repo": "bartowski/HuggingFaceTB_SmolLM3-3B-GGUF", "file": "HuggingFaceTB_SmolLM3-3B-Q4_K_M.gguf", "url": "https://huggingface.co/bartowski/HuggingFaceTB_SmolLM3-3B-GGUF/blob/main/HuggingFaceTB_SmolLM3-3B-Q4_K_M.gguf", "recommended": true }, { "id": "Q5_K_M", "bytes": 2213756736, "repo": "bartowski/HuggingFaceTB_SmolLM3-3B-GGUF", "file": "HuggingFaceTB_SmolLM3-3B-Q5_K_M.gguf", "url": "https://huggingface.co/bartowski/HuggingFaceTB_SmolLM3-3B-GGUF/blob/main/HuggingFaceTB_SmolLM3-3B-Q5_K_M.gguf", "recommended": false }, { "id": "Q6_K", "bytes": 2530860864, "repo": "bartowski/HuggingFaceTB_SmolLM3-3B-GGUF", "file": "HuggingFaceTB_SmolLM3-3B-Q6_K.gguf", "url": "https://huggingface.co/bartowski/HuggingFaceTB_SmolLM3-3B-GGUF/blob/main/HuggingFaceTB_SmolLM3-3B-Q6_K.gguf", "recommended": false }, { "id": "Q8_0", "bytes": 3275575104, "repo": "bartowski/HuggingFaceTB_SmolLM3-3B-GGUF", "file": "HuggingFaceTB_SmolLM3-3B-Q8_0.gguf", "url": "https://huggingface.co/bartowski/HuggingFaceTB_SmolLM3-3B-GGUF/blob/main/HuggingFaceTB_SmolLM3-3B-Q8_0.gguf", "recommended": false } ], "arch": { "kind": "dense", "layers": 36, "fullLayers": 36, "hidden": 2048, "heads": 16, "kvHeads": 4, "headDim": 128, "intermediate": 11008, "vocab": 128256, "linear": null, "experts": null, "expertsPerTok": null, "moeIntermediate": null, "sharedExpertIntermediate": null, "globalHeadDim": null, "globalKvHeads": null, "slidingWindow": null, "slidingLayers": 0, "tieEmbeddings": false }, "ollama": null, "multimodal": false } ] }