Model hub — Distillio

← Home

Hand-picked lab flagship models for fine-tuning (safetensors), and ready-to-run GGUF quantizations for local inference (Ollama, LM Studio, llama.cpp). Download links go to the original Hugging Face repos.

Fine-tune models (safetensors)

• Qwen/Qwen3.5-2B — transformers · qwen3_5 · image-text-to-text (2B params, 4,548,221,488, updated 2026-03-02)
• Qwen/Qwen3.5-4B — transformers · qwen3_5 · image-text-to-text (5B params, 9,319,828,096, updated 2026-03-02)
• Qwen/Qwen3-VL-8B-Instruct — transformers · qwen3_vl · image-text-to-text (9B params, 17,534,339,512, updated 2025-10-15)
• Qwen/Qwen3-4B-Instruct-2507 — transformers · qwen3 · conversational (4B params, 8,044,982,000, updated 2025-09-17)
• Qwen/Qwen3-8B — transformers · qwen3 · conversational (8B params, 16,381,516,776, updated 2025-07-26)
• Qwen/Qwen2.5-7B-Instruct — transformers · qwen2 · chat (8B params, 15,231,271,888, updated 2025-01-12)
• Qwen/Qwen2.5-Coder-7B-Instruct — transformers · qwen2 · code (8B params, 15,231,271,864, updated 2025-01-12)
• ornith-ai/Ornith-1.5-9B — transformers · qwen3_5 · image-text-to-text (10B params, 19,306,303,864, updated 2026-08-23)
• Pluto-AI-Labs/Apollo-VL-Edge-3B — qwen2_5_vl · vision (4B params, 7,509,337,112, updated 2026-08-16)
• Pluto-AI-Labs/Atlas-Frontier-Distill-3B — transformers · qwen2 · distillation (3B params, 6,650,999,224, updated 2026-08-03)
• HuggingFaceTB/SmolLM3-3B — transformers · smollm3 · conversational (3B params, 6,150,235,008, updated 2025-09-10)
• ibm-granite/granite-4.2-3b — transformers · granite · granite-4.2 (4B params, 7,319,517,120, updated 2026-10-20)
• ibm-granite/granite-4.2-8b — transformers · granite · granite-4.2 (9B params, 17,583,228,032, updated 2026-10-20)
• mistralai/Ministral-3-3B-Instruct-2512 — vllm · mistral3 (4B params, 9,344,188,292, updated 2026-07-15)
• mistralai/Ministral-3-8B-Instruct-2512 — vllm · mistral3 (9B params, 20,841,288,000, updated 2026-07-15)
• mistralai/Ministral-3-14B-Instruct-2512 — vllm · mistral3 (14B params, 31,453,437,040, updated 2026-07-15)
• microsoft/Fara1.5-4B — transformers · qwen3_5 · image-text-to-text (5B params, 9,078,620,456, updated 2026-07-27)
• openai/gpt-oss-20b — transformers · gpt_oss · vllm (21B params, 27,522,617,888, updated 2025-08-26)
• XiaomiMiMo/MiMo-VL-7B-RL-2508 — transformers · qwen2_5_vl · image-text-to-text (8B params, 16,612,526,808, updated 2026-10-28)
• tencent/Hy-MT2-1.8B — transformers · hunyuan_v1_dense · translation (2B params, 4,077,072,784, updated 2026-05-26)
• tencent/Hy-MT2-7B — transformers · hunyuan_v1_dense · translation (8B params, 16,059,122,256, updated 2026-05-26)
• tencent/Hy-MT2-30B-A3B — transformers · hy_v3 · translation (30B params, 60,131,744,832, updated 2026-07-21)
• tencent/Hunyuan-A13B-Instruct — transformers · hunyuan_v1_moe · conversational (80B params, 160,803,954,720, updated 2025-08-21)
• XiaomiMiMo/MiMo-7B-RL — transformers · mimo · conversational (8B params, 15,666,870,936, updated 2026-10-28)
• XiaomiMiMo/MiMo-V2.5 — transformers · mimo_v2 · multimodal (311B params, 315,693,004,496, updated 2026-10-28)
• Qwen/Qwen3.8-27B — transformers · qwen3_5 · image-text-to-text (27B params, 55,563,006,776, updated 2026-08-14)
• meta-models/Muse-Glimmer-30B — transformers · muse_glimmer · image-text-to-text (30B params, 59,553,435,272, updated 2026-10-22)
• google/gemma-4-12B — transformers · gemma4_unified · image-text-to-text (12B params, 23,919,549,408, updated 2026-10-30)
• google/gemma-4-32B — multimodal MoE (32B params)
• Qwen/Qwen3.6-35B-A3B — transformers · qwen3_5_moe · image-text-to-text (35B params, 71,903,776,776, updated 2026-04-24)
• Qwen/Qwen3.6-27B — transformers · qwen3_5 · image-text-to-text (27B params, 55,563,006,400, updated 2026-04-24)
• Qwen/Qwen3.6-27B-FP8 — transformers · qwen3_5 · image-text-to-text (27B params, 30,866,866,928, updated 2026-04-24)
• Qwen/Qwen3.5-9B — transformers · qwen3_5 · image-text-to-text (9B params, 19,306,310,880, updated 2026-03-02)
• microsoft/phi-4 — transformers · phi3 · phi (14B params, 29,319,042,992, updated 2026-07-14)
• zai-org/GLM-4.7-Flash — transformers · glm4_moe_lite · conversational (30B params, 62,444,175,504, updated 2026-01-29)
• deepseek-ai/DeepSeek-Coder-V2-Lite-Instruct — transformers · deepseek_v2 · conversational (16B params, 31,413,626,609, updated 2024-07-03)
• meta-llama/Llama-Guard-4-12B — transformers · llama4 · image-text-to-text (12B params, 24,002,322,024, updated 2025-04-29)
• NousResearch/Hermes-4.3-36B — transformers · seed_oss · Bytedance Seed (36B params, 72,302,298,208, updated 2025-12-06)
• NousResearch/Hermes-4-14B — transformers · qwen3 · Qwen-3-14B (14B params, 29,536,665,656, updated 2026-01-09)

Local GGUF models

Quantized for Ollama / LM Studio / llama.cpp, ordered by download popularity. Click a repo link to see its exact quant files on Hugging Face.

• ornith-ai/Ornith-1.5-9B-GGUF — transformers · conversational (9B params, 5,426,513 downloads, updated 2026-08-24)
• ornith-ai/Ornith-1.0-9B-GGUF — transformers · conversational (9B params, 2,414,835 downloads, updated 2026-06-25)
• unsloth/gemma-4-12b-it-GGUF — gemma4 · gemma (12B params, 2,023,832 downloads, updated 2026-07-17)
• unsloth/Qwen3.5-9B-GGUF — transformers · image-text-to-text · base_model:Qwen/Qwen3.5-9B (9B params, 1,554,745 downloads, updated 2026-03-02)
• Qwen/Qwen3-VL-8B-Instruct-GGUF — transformers · image-text-to-text · base_model:Qwen/Qwen3-VL-8B-Instruct (8B params, 1,406,262 downloads, updated 2025-11-01)
• unsloth/Qwen3.5-4B-GGUF — transformers · image-text-to-text · base_model:Qwen/Qwen3.5-4B (4B params, 1,323,712 downloads, updated 2026-03-02)
• tencent/Hy-MT2-7B-GGUF — base_model:tencent/Hy-MT2-7B · base_model:quantized:tencent/Hy-MT2-7B (7B params, 1,057,596 downloads, updated 2026-05-26)
• bartowski/MiMo-V2.6-Distill-Qwen-9B-GGUF — mimo_v2 · agentic (9B params, 974,177 downloads, updated 2026-09-21)
• Qwen/Qwen3-8B-GGUF — base_model:Qwen/Qwen3-8B · base_model:quantized:Qwen/Qwen3-8B (8B params, 733,491 downloads, updated 2025-05-21)
• Qwen/Qwen3-VL-4B-Instruct-GGUF — transformers · image-text-to-text · base_model:Qwen/Qwen3-VL-4B-Instruct (4B params, 638,421 downloads, updated 2025-11-01)
• Qwen/Qwen3-4B-GGUF — base_model:Qwen/Qwen3-4B · base_model:quantized:Qwen/Qwen3-4B (4B params, 617,066 downloads, updated 2025-05-21)
• unsloth/Qwen3-4B-Instruct-2507-GGUF — transformers · qwen · qwen3 (4B params, 586,058 downloads, updated 2025-08-20)
• unsloth/gemma-4-E4B-it-GGUF — gemma4 · gemma (4B params, 571,394 downloads, updated 2026-07-17)
• unsloth/gemma-4-E2B-it-GGUF — gemma4 · gemma (2B params, 532,322 downloads, updated 2026-07-17)
• unsloth/Qwen3-4B-GGUF — transformers · qwen3 · qwen (4B params, 531,614 downloads, updated 2025-06-08)
• unsloth/Qwen3.5-2B-GGUF — transformers · image-text-to-text · base_model:Qwen/Qwen3.5-2B (2B params, 426,284 downloads, updated 2026-03-02)
• unsloth/Qwen3-VL-4B-Instruct-GGUF — transformers · qwen · qwen3 (4B params, 400,864 downloads, updated 2025-10-31)
• unsloth/Ornith-1.0-9B-GGUF — qwen3_5_moe · base_model:ornith-ai/Ornith-1.0-9B (9B params, 388,215 downloads, updated 2026-07-18)
• bartowski/Meta-Llama-3.1-8B-Instruct-GGUF — facebook · meta (8B params, 366,340 downloads, updated 2024-12-01)
• Qwen/Qwen3-14B-GGUF — base_model:Qwen/Qwen3-14B · base_model:quantized:Qwen/Qwen3-14B (14B params, 353,321 downloads, updated 2025-05-09)
• tencent/Hy-MT2-1.8B-GGUF — base_model:tencent/Hy-MT2-1.8B · base_model:quantized:tencent/Hy-MT2-1.8B (1.8B params, 325,655 downloads, updated 2026-09-08)
• Qwen/Qwen2.5-Coder-7B-Instruct-GGUF — transformers · code · codeqwen (7B params, 325,382 downloads, updated 2024-11-12)
• unsloth/Qwen3.5-0.8B-GGUF — transformers · image-text-to-text · base_model:Qwen/Qwen3.5-0.8B (0.8B params, 284,620 downloads, updated 2026-03-02)
• unsloth/NVIDIA-Nemotron-3.5-Bolt-E4B-GGUF — nvidia · nemotron-3.5 (4B params, 281,538 downloads, updated 2026-10-24)
• ibm-granite/granite-4.2-3b-GGUF — transformers · language · granite-4.2 (3B params, 269,920 downloads, updated 2026-09-02)
• ibm-granite/granite-4.2-8b-GGUF — llama.cpp · language (8B params, 268,281 downloads, updated 2026-09-02)
• unsloth/Qwen3-1.7B-GGUF — transformers · qwen3 · qwen (1.7B params, 242,630 downloads, updated 2025-06-08)
• unsloth/Qwen2.5-VL-7B-Instruct-GGUF — transformers · qwen2_5_vl · image-text-to-text (7B params, 240,086 downloads, updated 2025-05-12)
• Qwen/Qwen2.5-Coder-1.5B-Instruct-GGUF — transformers · code · codeqwen (1.5B params, 237,373 downloads, updated 2024-11-12)
• Qwen/Qwen3-0.6B-GGUF — base_model:Qwen/Qwen3-0.6B · base_model:quantized:Qwen/Qwen3-0.6B (0.6B params, 222,170 downloads, updated 2025-05-09)
• unsloth/Qwen3-VL-8B-Instruct-GGUF — transformers · image-text-to-text · base_model:Qwen/Qwen3-VL-8B-Instruct (8B params, 215,179 downloads, updated 2025-10-31)
• bartowski/Qwen_Qwen3.5-9B-GGUF — image-text-to-text · base_model:Qwen/Qwen3.5-9B (9B params, 199,826 downloads, updated 2026-05-19)
• Qwen/Qwen3-VL-2B-Instruct-GGUF — transformers · image-text-to-text · base_model:Qwen/Qwen3-VL-2B-Instruct (2B params, 197,323 downloads, updated 2025-11-01)
• Qwen/Qwen2.5-Coder-14B-Instruct-GGUF — transformers · code · codeqwen (14B params, 193,090 downloads, updated 2024-11-12)
• Qwen/Qwen3-Embedding-0.6B-GGUF — base_model:Qwen/Qwen3-0.6B-Base · base_model:quantized:Qwen/Qwen3-0.6B-Base (0.6B params, 167,454 downloads, updated 2025-07-14)
• Qwen/Qwen3-1.7B-GGUF — base_model:Qwen/Qwen3-1.7B · base_model:quantized:Qwen/Qwen3-1.7B (1.7B params, 157,743 downloads, updated 2025-05-09)
• bartowski/DeepSeek-R1-Distill-Qwen-7B-GGUF — base_model:deepseek-ai/DeepSeek-R1-Distill-Qwen-7B · base_model:quantized:deepseek-ai/DeepSeek-R1-Distill-Qwen-7B (7B params, 136,516 downloads, updated 2025-03-07)
• bartowski/Qwen_Qwen3-0.6B-GGUF — base_model:Qwen/Qwen3-0.6B · base_model:quantized:Qwen/Qwen3-0.6B (0.6B params, 128,028 downloads, updated 2025-04-28)
• bartowski/Qwen_Qwen3.5-4B-GGUF — image-text-to-text · base_model:Qwen/Qwen3.5-4B (4B params, 122,834 downloads, updated 2026-05-19)
• bartowski/Qwen_Qwen3-4B-GGUF — base_model:Qwen/Qwen3-4B · base_model:quantized:Qwen/Qwen3-4B (4B params, 105,951 downloads, updated 2025-04-28)