🔮All Our Models

See the table below for all Dynamic GGUF, 4-bit, 16-bit uploaded models on Hugging Face.

• GGUFs can be used to run in your favorite places like Ollama, Open WebUI and llama.cpp.

• 4-bit and 16-bit models can be used for inference serving or fine-tuning.

• GGUF + 4-bit

Here's a table of all our GGUF + 4-bit model uploads:

Model	GGUF	Instruct (4-bit)	Base (4-bit)
DeepSeek-R1	R1-0528 - new R1-0528-Qwen3-8B - new R1 R1 Zero Llama 3 (8B) Llama 3.3 (70B) Qwen 2.5 (14B) Qwen 2.5 (32B) Qwen 2.5 (1.5B) Qwen 2.5 (7B)	R1-0528-Qwen3-8B - new Llama 3 (8B) Llama 3.3 (70B) Qwen 2.5 (14B) Qwen 2.5 (32B) Qwen 2.5 (1.5B) Qwen 2.5 (7B)
Qwen3 (new)	0.6B 1.7B 4B 8B 14B 30B-A3B 32B 235B-A22B	0.6B 1.7B 4B 8B 14B 30B-A3B 32B	0.6B 1.7B 4B 8B 14B 30B
Llama 4	Scout Maverick	Scout	Scout
Gemma 3	1B 4B 12B 27B	1B 4B 12B 27B	1B 4B 12B 27B
Mistral	Magistral - new Small 3.1 Devstral - new Small 3 NeMo 2407 (12B)	Magistral - new Small 3.1 Devstral - new Small 3 NeMo 2407 (12B) Small 2409 (22B) Large 2407 7B (v0.3) 7B (v0.2) Pixtral (12B) 2409 Mixtral-8x7B	Small 3.1 Small 3 NeMo 2407 (12B) 7B (v0.3) 7B (v0.2) Pixtral (12B) 2409
Qwen2.5-Omni (new)	3B 7B
Llama 3.2	1B 3B	1B 3B 11B Vision 90B Vision	1B 3B 11B Vision 90B Vision
Phi-4	Reasoning-plus Reasoning Mini-reasoning Phi-4 mini	Reasoning-plus Reasoning Mini-reasoning Phi-4 mini
Text-to-speech (TTS)		Orpheus-3B Original models: Sesame-CSM (1B) Whisper Large V3 - STT Llasa-TTS (1B) Spark-TTS (0.5B) Oute-TTS (1B)	Orpheus-3B
Llama 3.3	70B	70B
Llama 3.1	8B	8B 70B 405B	8B 70B 405B
DeepSeek V3	V3-0324 V3
Qwen2.5-VL	3B 7B 32B 72B	3B 7B 32B 72B
Qwen 2.5		0.5B 1.5B 3B 7B 14B 32B 72B QwQ QVQ	0.5B 1.5B 3B 7B 14B 32B 72B
QwQ-32B	32B	32B
Gemma 2	All variants	2B 9B 27B	2B 9B 27B
Phi-3.5		mini
Phi-3		mini medium
Llama 3		8B 70B	8B 70B
Llava		1.5 (7B) 1.6 Mistral (7B)
Qwen 2.5 Coder	0.5B 1.5B 3B 7B 14B 32B	0.5B 1.5B 3B 7B 14B 32B	0.5B 1.5B 3B 7B 14B 32B
Llama 2		7B	7B 13B
Qwen2 VL		2B 7B 72B
SmolLM2	135M 360M 1.7B	135M 360M 1.7B	135M 360M 1.7B
TinyLlama		Instruct	Base
Qwen2		1.5B 7B 72B	1.5B 7B 72B
Zephyr SFT		Instruct
CodeLlama			7B 13B 34B
Yi		34B	6B (v 1.5) 6B 34B

• 16-bit original

Here's a table of all our 16-bit or 8-bit original model uploads:

Model	Instruct	Base
Qwen3 (new)	0.6B 1.7B 4B 8B 14B 30B-A3B 32B 235B-A22B	0.6B 1.7B 4B 8B 14B 30B-A3B
Llama 4 (new)	Scout Maverick	Scout Maverick
Phi-4 (new)	Reasoning-plus Reasoning ' Phi-4 Phi-4-mini
Gemma 3 (new)	1B 4B 12B 27B	1B 4B 12B 27B
DeepSeek-R1	R1 R1 Zero Llama 3 (8B) Llama 3.3 (70B) Qwen 2.5 (14B) Qwen 2.5 (32B) Qwen 2.5 (1.5B) Qwen 2.5 (7B) R1 (bf16)
Llama 3.2	1B 3B 11B Vision 90B Vision	1B 3B 11B Vision 90B Vision
Llama 3.3	70B
Llama 3.1	8B 70B	8B 70B
Mistral	Mistral Small 2501 NeMo 2407 (12B) Small 2409 (22B) 7B (v0.3) 7B (v0.2) Pixtral (12B) 2409 Mixtral-8x7B	Mistral Small 2501 NeMo 2407 (12B) 7B (v0.3) 7B (v0.2) Pixtral (12B) 2409
Gemma 2	2B 9B 27B	2B 9B 27B
DeepSeek V3	bf16 original 8-bit
Phi-3.5	mini
Phi-3	mini medium
Llama 3	8B	8B
Qwen 2.5	0.5B 1.5B 3B 7B 14B 32B 72B	0.5B 1.5B 3B 7B 14B 32B 72B
Llava	1.5 (7B) 1.6 Mistral (7B)
Qwen2 VL	2B 7B 72B
Llama 2	7B	7B 13B
SmolLM2	135M 360M 1.7B	135M 360M 1.7B
TinyLlama	Instruct	Base
Qwen2	1.5B 7B	1.5B 7B
Zephyr SFT	Instruct