unsloth/gemma-3-4b-it-GGUF
Image-Text-to-Text • 4B • Updated • 175k
• 217
Note Good, supports images
unsloth/Qwen3-4B-Instruct-2507-GGUF
4B • Updated • 884k
• 227
Note Also has thinking variant
NousResearch/Hermes-3-Llama-3.2-3B-GGUF
3B • Updated • 6.53k
• 76
Note Great for simple tasks, Llama 3.2 also has 1B variant
bartowski/microsoft_Phi-4-mini-instruct-GGUF
Text Generation • 4B • Updated • 90.3k
• 52
Note Great for certain usecases, has a reasoning variant
Image-Text-to-Text • 2B • Updated • 2.37M
• 1.45k
Note All in one solution for vision
ibm-granite/granite-docling-258M
Image-Text-to-Text • 0.3B • Updated • 380k
• 1.3k
Note Document OCR
Text Generation • 1B • Updated • 12.1k
• 111
Note New arch
unsloth/orpheus-3b-0.1-ft-GGUF
Text-to-Speech • 3B • Updated • 14.7k
• 43
Note Good TTS, supports emotion tags
nvidia/parakeet-tdt-0.6b-v2
Automatic Speech Recognition • Updated • 130k
• 1.59k
Note Great STT, v3 is multilingual
Text Generation • 4B • Updated • 5.14k
• 75
Note Deepsearch type of usecases
Text-to-Speech • Updated • 13M
• • 7.45k
Note Great TTS
unsloth/Qwen3-VL-4B-Thinking-GGUF
Image-Text-to-Text • 4B • Updated • 7.29k
• 27