{ "schema_version": 3, "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": { "chat-template-kwargs": "{\"reasoning_effort\":\"medium\"}", "temp": "1.0", "top-p": "0.95", "top-k": "20", "min-p": "0.0" }, "quality": 90, "decode_fraction": 1.0, "draft": { "path": "Qwen3.8-27B-DFlash2-Q4_K_M.gguf", "size_bytes": 1143006816, "repo": "z-lab/Qwen3.8-27B-DFlash2-GGUF", "revision": "2d9571f8ce46e151f61c6499c99dee6079e1d610" }, "draft_spec_type": "draft-dflash", "draft_depth": 7, "draft_min_engine": "b11370" }, { "id": "qwen3.8-flash-next", "display_name": "Qwen3.8 Flash Next", "description": "Frontier-scale mixture-of-experts with multi-token prediction; sees images", "repo": "unsloth/Qwen3.8-Flash-Next-GGUF", "variants": [ { "quant": "UD-IQ4_XS", "files": [ { "path": "UD-IQ4_XS/Qwen3.8-Flash-Next-UD-IQ4_XS-00001-of-00003.gguf", "size_bytes": 10946624, "revision": "38bb39ee97821de2c9009abb7e93950eec396e66" }, { "path": "UD-IQ4_XS/Qwen3.8-Flash-Next-UD-IQ4_XS-00002-of-00003.gguf", "size_bytes": 49835229856, "revision": "38bb39ee97821de2c9009abb7e93950eec396e66" }, { "path": "UD-IQ4_XS/Qwen3.8-Flash-Next-UD-IQ4_XS-00003-of-00003.gguf", "size_bytes": 43836407744, "revision": "38bb39ee97821de2c9009abb7e93950eec396e66" } ], "lazy_bytes": 28800138240 } ], "n_ctx_train": 262144, "full_layers": 12, "recurrent_layers": 36, "per_layer_f16": 2048, "moe": true, "n_vocab": 248320, "architecture": "qwen4exp", "mtp_draft_depth": 3, "mmproj": { "path": "mmproj-BF16.gguf", "size_bytes": 907542944, "local": "mmproj-Qwen3.8-Flash-Next-BF16.gguf", "revision": "38bb39ee97821de2c9009abb7e93950eec396e66" }, "mtp_head": { "repo": "ggml-org/Qwen3.8-Flash-Next-GGUF", "path": "mtp-Qwen3.8-Flash-Next-Q8_0.gguf", "size_bytes": 4137429280, "revision": "052beeaca7bec4a303e59cc7bc630c4f3a1b845d", "local": "mtp-Qwen3.8-Flash-Next-ggml-org-Q8_0.gguf" }, "min_engine": "b11370", "quality": 95, "decode_fraction": 0.08 }, { "id": "nemotron-3.5-bolt-74b-a9b", "display_name": "Nemotron 3.5 Bolt", "description": "NVIDIA's 74B mixture-of-experts in a 2-bit build; sees images", "repo": "nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB-GGUF", "variants": [ { "quant": "IQ2", "files": [ { "path": "NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB.gguf", "size_bytes": 22018326528, "revision": "e0fecb69cb3e5e4c5e8a056fa2aec98c78e91246" } ] } ], "n_ctx_train": 262144, "full_layers": 8, "recurrent_layers": 80, "per_layer_f16": 1024, "moe": true, "n_vocab": 131072, "architecture": "nemotron_h_moe", "mmproj": { "path": "mmproj-NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB.gguf", "size_bytes": 1733692736, "revision": "e0fecb69cb3e5e4c5e8a056fa2aec98c78e91246" }, "min_engine": "b11671", "decode_fraction": 0.22, "recommend": false, "draft": { "path": "NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB-LiLiCorr.gguf", "size_bytes": 998242560, "repo": "nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB-LiLiCorr-GGUF", "revision": "182d2987939e12f458298b5d5f6ef98912f4ad6c" }, "draft_spec_type": "draft-dflash", "draft_depth": 7, "draft_min_engine": "b11671" }, { "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 } ] }