AImodelsComparison.com 1,011 models
Menu
Models Best of Trending Can I run it? Methodology About
1,011 models · refreshed nightly

Models that run on RTX 4070 · 16 GB

Every model in the catalog with its licence, estimated VRAM and daily-tracked downloads. Filters update the URL — share any view.

Task
All text-generation 400 text-to-image 150 ASR 89 image-text-to-text 82 sentence-similarity 55 feature-extraction 47 translation 44 text-classification 39 image-classification 32 token-classification 22 text-to-speech 16 audio-classification 12 image-segmentation 8 object-detection 8 summarization 7
Runs on
Any RTX 3060 · 12 GB RTX 4070 · 16 GB RTX 4090 · 24 GB 2 × 24 GB Mac M3 · 24 GB Mac M4 Max · 64 GB
Licence
Size
$
542 models
#ModelParamsContextCommercial use30dOn RTX 4070 · 16 GB
101 bge-base-en-v1.5-course-recommender-v5 datasocietyco · sentence-similarity 110 M 512 unknown 1.8 M 1.0 GB
102 ko-sroberta-multitask jhgan · sentence-similarity 110 M 512 unknown 1.8 M 1.0 GB
103 Qwen2-VL-7B-Instruct-AWQ Qwen · image-text-to-text 8.3 B 33 K ✓ apache-2.0 1.8 M 9.4 GB
104 Llama-3.2-1B meta-llama · text-generation · gated 1.2 B — ⚠ llama3.2 1.8 M 3.4 GB
105 wav2vec2-base-960h facebook · ASR 90 M — ✓ apache-2.0 1.7 M 0.9 GB
106 stsb-bert-tiny-safetensors sentence-transformers-testing · sentence-similarity 0 M 512 unknown 1.7 M 0.5 GB
107 gte-large-en-v1.5 Alibaba-NLP · sentence-similarity 430 M 8 K ✓ apache-2.0 1.7 M 2.5 GB
108 distil-large-v3 distil-whisper · ASR 760 M — ✓ mit 1.7 M 5.6 GB
109 tf_efficientnetv2_s.in21k_ft_in1k timm · image-classification 20 M — ✓ apache-2.0 1.6 M 0.6 GB
110 Qwen3.6-27B-Fable-Fusion-711-Uncensored-Heretic-NM-DAU-NEO-MAX-MTP-GGUF DavidAU · image-text-to-text — — ✓ apache-2.0 1.6 M 13.8 GB
111 t5-base google-t5 · translation 220 M — ✓ apache-2.0 1.6 M 1.5 GB
112 SapBERT-from-PubMedBERT-fulltext cambridgeltl · feature-extraction 110 M 512 ✓ apache-2.0 1.6 M 1.0 GB
113 paraphrase-mpnet-base-v2 sentence-transformers · sentence-similarity 110 M 512 ✓ apache-2.0 1.6 M 1.0 GB
114 Qwen3-TTS-12Hz-0.6B-CustomVoice Qwen · text-to-speech 910 M — ✓ apache-2.0 1.6 M 3.4 GB
115 Qwen3-0.6B-FP8 Qwen · text-generation 750 M 41 K ✓ apache-2.0 1.6 M 1.8 GB
116 bge-small-en BAAI · feature-extraction 30 M 512 ✓ mit 1.6 M 0.7 GB
117 bart-large-cnn facebook · summarization 410 M 1 K ✓ mit 1.6 M 2.3 GB
118 gender-classification rizvandwiki · image-classification 90 M — unknown 1.5 M 0.9 GB
119 nomic-embed-text-v2-moe nomic-ai · sentence-similarity 480 M — ✓ apache-2.0 1.5 M 2.7 GB
120 OpenELM-1_1B-Instruct apple · text-generation 1.1 B — apple-amlr 1.5 M 3.0 GB
121 SmolLM-1.7B-Instruct-quantized.w4a16 nm-testing · text-generation 1.8 B 2 K ✓ apache-2.0 1.5 M 2.7 GB
122 Llama-3.2-1B-Instruct-FP8-dynamic RedHatAI · text-generation 1.5 B 131 K ⚠ llama3.2 1.5 M 3.0 GB
123 gpt2-large openai-community · text-generation 810 M — ✓ mit 1.5 M 4.2 GB
124 table-transformer-structure-recognition microsoft · object-detection 30 M 1 K ✓ mit 1.5 M 0.6 GB
125 Llama-3.2-3B-Instruct meta-llama · text-generation · gated 3.2 B — ⚠ llama3.2 1.5 M 8.0 GB
126 InternVL2-2B OpenGVLab · image-text-to-text 2.2 B — ✓ mit 1.5 M 5.7 GB
127 parakeet-tdt-0.6b-v3 mlx-community · ASR 630 M — ✓ cc-by-4.0 1.4 M 3.4 GB
128 snowflake-arctic-embed-l-v2.0 Snowflake · sentence-similarity 570 M 8 K ✓ apache-2.0 1.4 M 3.1 GB
129 gemma-4-26B-A4B-it-GGUF unsloth · image-text-to-text — — ✓ apache-2.0 1.4 M 11.4 GB
130 wav2vec2-xls-r-300m-cs-250 comodoro · ASR 320 M — ✓ apache-2.0 1.4 M 1.9 GB
131 w2v-bert-2.0 facebook · feature-extraction 580 M — ✓ mit 1.4 M 3.1 GB
132 Qwen3-VL-Embedding-2B Qwen · sentence-similarity 2.1 B — ✓ apache-2.0 1.4 M 5.5 GB
133 Qwen3.5-9B-AWQ QuantTrio · image-text-to-text 9.7 B — ✓ apache-2.0 1.3 M 15.6 GB
134 Phi-3.5-vision-instruct microsoft · image-text-to-text 4.2 B 131 K ✓ mit 1.3 M 10.2 GB
135 bert-base-NER dslim · token-classification 110 M 512 ✓ mit 1.3 M 1.0 GB
136 h2ovl-mississippi-800m h2oai · text-generation 830 M — ✓ apache-2.0 1.3 M 2.4 GB
137 efficientnet_b0.ra_in1k timm · image-classification 10 M — ✓ apache-2.0 1.3 M 0.5 GB
138 h2ovl-mississippi-2b h2oai · text-generation 2.2 B — ✓ apache-2.0 1.3 M 5.6 GB
139 deepseek-vl2-tiny deepseek-ai · image-text-to-text 3.4 B — other 1.3 M 8.4 GB
140 gte-multilingual-base Alibaba-NLP · sentence-similarity 310 M 8 K ✓ apache-2.0 1.2 M 1.2 GB
141 romanian-wav2vec2 gigant · ASR 320 M — ✓ apache-2.0 1.2 M 1.9 GB
142 table-transformer-detection microsoft · object-detection 30 M 1 K ✓ mit 1.2 M 0.6 GB
143 Qwen3.6-27B-MTP-GGUF unsloth · image-text-to-text — — ✓ apache-2.0 1.2 M 14.3 GB
144 ast-finetuned-audioset-10-10-0.4593 MIT · audio-classification 90 M — ✓ bsd-3-clause 1.2 M 0.9 GB
145 Qwen3.5-4B-GGUF unsloth · image-text-to-text — — ✓ apache-2.0 1.2 M 2.8 GB
146 gte-Qwen2-1.5B-instruct Alibaba-NLP · sentence-similarity 1.8 B 131 K ✓ apache-2.0 1.2 M 8.6 GB
147 bloomz-560m bigscience · text-generation 560 M — ⚠ bigscience-bloom-rail-1.0 1.2 M 1.8 GB
148 resnet-50 microsoft · image-classification 30 M — ✓ apache-2.0 1.2 M 0.6 GB
149 Qwen3-4B-Instruct-2507-FP8 Qwen · text-generation 4.4 B 262 K ✓ apache-2.0 1.2 M 6.9 GB
150 SmolVLM2-500M-Video-Instruct HuggingFaceTB · image-text-to-text 510 M — ✓ apache-2.0 1.2 M 2.8 GB
VRAM figures are estimates for the smallest available quantization at 8K context — see /methodology. Downloads refresh nightly from the Hugging Face API.