{ "schema_version": 1, "models": [ { "id": "qwen3.8-27b", "display_name": "Qwen3.8 27B", "description": "Best all-round agent model; sees images; long context stays fast", "repo": "unsloth/Qwen3.8-27B-GGUF", "variants": [ { "quant": "UD-Q4_K_M", "files": [ { "path": "Qwen3.8-27B-UD-Q4_K_M.gguf", "size_bytes": 16464440224 } ] } ], "n_ctx_train": 262144, "full_layers": 16, "recurrent_layers": 48, "per_layer_f16": 4096, "n_vocab": 248320, "mmproj": { "path": "mmproj-BF16.gguf", "size_bytes": 931146432, "local": "mmproj-Qwen3.8-27B-BF16.gguf" }, "mtp": true, "mtp_draft_depth": 2, "sampling": { "temp": "1.0", "top-p": "0.95", "top-k": "20", "min-p": "0.0" }, "quality": 90, "decode_fraction": 1.0 }, { "id": "qwen3.8-flash-next", "display_name": "Qwen3.8 Flash Next", "description": "Frontier-scale model; needs a very large GPU to run well", "repo": "unsloth/Qwen3.8-Flash-Next-GGUF", "variants": [ { "quant": "UD-Q4_K_XL", "files": [ { "path": "UD-Q4_K_XL/Qwen3.8-Flash-Next-UD-Q4_K_XL-00001-of-00004.gguf", "size_bytes": 10946624 }, { "path": "UD-Q4_K_XL/Qwen3.8-Flash-Next-UD-Q4_K_XL-00002-of-00004.gguf", "size_bytes": 49859583136 }, { "path": "UD-Q4_K_XL/Qwen3.8-Flash-Next-UD-Q4_K_XL-00003-of-00004.gguf", "size_bytes": 49376141504 }, { "path": "UD-Q4_K_XL/Qwen3.8-Flash-Next-UD-Q4_K_XL-00004-of-00004.gguf", "size_bytes": 12087983520 } ] } ], "n_ctx_train": 262144, "full_layers": 12, "recurrent_layers": 36, "per_layer_f16": 2048, "moe": true, "n_vocab": 248320, "mmproj": { "path": "mmproj-BF16.gguf", "size_bytes": 907542944, "local": "mmproj-Qwen3.8-Flash-Next-BF16.gguf" }, "min_engine": "b10678", "quality": 95, "decode_fraction": 0.08 }, { "id": "qwen3.6-35b-a3b", "display_name": "Qwen3.6 35B-A3B", "description": "Bigger mixture-of-experts with multi-token prediction; sees images", "repo": "unsloth/Qwen3.6-35B-A3B-MTP-GGUF", "variants": [ { "quant": "UD-Q4_K_M", "files": [ { "path": "Qwen3.6-35B-A3B-UD-Q4_K_M.gguf", "size_bytes": 22663387424 } ], "validated": true } ], "n_ctx_train": 262144, "full_layers": 10, "recurrent_layers": 30, "per_layer_f16": 2048, "moe": true, "mtp": true, "n_vocab": 248320, "mtp_draft_depth": 2, "mmproj": { "path": "mmproj-BF16.gguf", "size_bytes": 902822528, "local": "mmproj-Qwen3.6-35B-A3B-BF16.gguf" }, "sampling": { "temp": "1.0", "top-p": "0.95", "top-k": "20", "min-p": "0.0" }, "quality": 80, "decode_fraction": 0.15 }, { "id": "deepseek-v4-flash", "display_name": "DeepSeek V4 Flash", "description": "Frontier-class model for machines with 128GB+ memory", "repo": "unsloth/DeepSeek-V4-Flash-0731-GGUF", "variants": [ { "quant": "UD-Q4_K_XL", "files": [ { "path": "UD-Q4_K_XL/DeepSeek-V4-Flash-0731-UD-Q4_K_XL-00001-of-00005.gguf", "size_bytes": 5257408 }, { "path": "UD-Q4_K_XL/DeepSeek-V4-Flash-0731-UD-Q4_K_XL-00002-of-00005.gguf", "size_bytes": 48935523072 }, { "path": "UD-Q4_K_XL/DeepSeek-V4-Flash-0731-UD-Q4_K_XL-00003-of-00005.gguf", "size_bytes": 48980787136 }, { "path": "UD-Q4_K_XL/DeepSeek-V4-Flash-0731-UD-Q4_K_XL-00004-of-00005.gguf", "size_bytes": 49999168416 }, { "path": "UD-Q4_K_XL/DeepSeek-V4-Flash-0731-UD-Q4_K_XL-00005-of-00005.gguf", "size_bytes": 7174505088 } ] } ], "n_ctx_train": 1048576, "full_layers": 43, "recurrent_layers": 0, "per_layer_f16": 1152, "moe": true, "n_vocab": 163840, "draft": { "path": "dspark-DeepSeek-V4-Flash-0731-Q8_0.gguf", "size_bytes": 10896057440 }, "sampling": { "temp": "1.0", "top-p": "0.95", "min-p": "0.01" }, "quality": 85, "decode_fraction": 0.1 } ] }