AImodelsComparison.com 1,089 open · 210 API
Menu
Models Best of Trending Can I run it? Methodology MCP server About
1,423 models · 1,089 open weights · 210 API-only · refreshed nightly

All models

Every model in the catalog — open weights you can self-host and API-only models — with licence or API price, estimated VRAM and daily-tracked downloads. Filters update the URL — share any view.

Type
All 1,299 Open weights (self-host) 1,089 API-only 210
Task
All text-generation 735 text-to-image 165 image-text-to-text 110 ASR 81 sentence-similarity 63 translation 51 text-classification 50 feature-extraction 46 image-classification 42 text-to-speech 23 token-classification 22 summarization 9 image-segmentation 9 audio-classification 8 object-detection 7 audio-text-to-text 1 any-to-any 1
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
$
1,423 models
#ModelParamsContextCommercial use30dMin VRAM
1001 DialoGPT-medium microsoft · text-generation Open weights — — ✓ mit 140 K —
1002 Wan2.2-I2V_General-NSFW-LoRA lopi999 · text-to-image Open weights — — unknown 140 K from 1.8 GB
1003 tiny-mpt-random-remote-code echarlaix · text-generation Open weights — — ✓ apache-2.0 140 K —
1004 Qwen3-235B-A22B-Instruct-2507 Qwen · text-generation Open weights Also via API 235.1 B 262 K ✓ apache-2.0 139 K from 553.0 GB
1005 typhoon2.5-qwen3-4b typhoon-ai · text-generation Open weights 4.0 B 262 K ✓ apache-2.0 139 K from 10.0 GB
1006 opus-mt-es-en Helsinki-NLP · translation Open weights — 512 ✓ apache-2.0 139 K —
1007 DeepSeek-R1-Distill-Qwen-7B-GGUF bartowski · text-generation Open weights 7.0 B 131 K unknown 139 K from 4.6 GB
1008 JevK5 alibiserikbay · text-generation Open weights 4.2 B 262 K ✓ apache-2.0 138 K from 10.4 GB
1009 Qwen-Image-GGUF unsloth · text-to-image Open weights 20.4 B — ✓ apache-2.0 138 K from 12.1 GB
1010 MiMo-V2.5 XiaomiMiMo · text-generation Open weights Also via API 310.8 B 1.0 M ✓ mit 137 K from 394.4 GB
1011 Hy-Memory-30B-A3B-GGUF rapatsky · text-generation Open weights 30.0 B — ✓ apache-2.0 136 K from 25.1 GB
1012 Meta-Llama-3-70B meta-llama · text-generation · gated Open weights 70.6 B — ⚠ llama3 136 K from 166.3 GB
1013 Qwopus3.6-27B-Fusion-GGUF KyleHessling1 · text-generation Open weights — — other 136 K from 15.4 GB
1014 needle3 Cactus-Compute · text-generation Open weights — 8 K ✓ apache-2.0 135 K from 0.8 GB
1015 NVIDIA-Nemotron-3.5-Bolt-E4B-4GB nvidia · text-generation Open weights 4.0 B — other 135 K from 7.9 GB
1016 opus-mt-ru-en Helsinki-NLP · translation Open weights — 512 ✓ cc-by-4.0 133 K —
1017 GLM-4.7-Flash-MLX-8bit lmstudio-community · text-generation Open weights 29.9 B 203 K ✓ mit 132 K from 40.0 GB
1018 GLM-5.2-FP8 zai-org · text-generation Open weights 753.3 B 1.0 M ✓ mit 132 K from 944.7 GB
1019 Olmo-3-7B-Instruct allenai · text-generation Open weights 7.3 B 66 K ✓ apache-2.0 132 K from 17.7 GB
1020 Llama-3.2-1B-Instruct-GGUF bartowski · text-generation Open weights — — ⚠ llama3.2 132 K from 1.2 GB
1021 Meta-Llama-3.1-8B-Instruct NousResearch · text-generation Open weights 8.0 B 131 K ⚠ llama3.1 131 K from 19.4 GB
1022 ERNIE-Image Comfy-Org · text-to-image Open weights 8.0 B — ✓ apache-2.0 130 K from 53.5 GB
1023 opus-mt-it-en Helsinki-NLP · translation Open weights — 512 ✓ apache-2.0 128 K —
1024 opus-mt-en-es Helsinki-NLP · translation Open weights — 512 ✓ apache-2.0 127 K —
1025 MeinaMix Meina · text-to-image Open weights — — ⚠ creativeml-openrail-m 127 K from 68.4 GB
1026 FLUX.1-schnell unsloth · text-to-image Open weights 11.9 B — ✓ apache-2.0 124 K from 65.9 GB
1027 opus-mt-nl-en Helsinki-NLP · translation Open weights — 512 ✓ apache-2.0 123 K —
1028 Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16 nvidia · any-to-any Open weights Also via API 33.0 B — other 122 K from 78.1 GB
1029 Qwen-Image-2.1-GGUF leejet · text-to-image Open weights — — unknown 119 K from 3.3 GB
1030 lcm-lora-sdv1-5 latent-consistency · text-to-image Open weights — — ⚠ openrail++ 116 K from 0.6 GB
1031 Krea-2-Turbo-GGUF vantagewithai · text-to-image Open weights — — other 116 K from 5.9 GB
1032 CyberRealisticPony cyberdelia · text-to-image Open weights — — ⚠ creativeml-openrail-m 114 K from 298.0 GB
1033 Qwen-Image-GGUF QuantStack · text-to-image Open weights — — ✓ apache-2.0 113 K from 0.8 GB
1034 NetaYume-Lumina-Image-2.0 duongve · text-to-image Open weights — — ✓ apache-2.0 112 K from 99.5 GB
1035 Flux2-Klein-9B-True-V2 wikeeyang · text-to-image Open weights — — other 112 K from 6.8 GB
1036 GLM-4.5 zai-org · text-generation Open weights Also via API 358.3 B 131 K ✓ mit 111 K from 842.6 GB
1037 Llama-4-Scout-17B-16E-Instruct meta-llama · image-text-to-text · gated Open weights Also via API 108.6 B — other 110 K from 255.8 GB
1038 z-image-turbo-nsfw-v2-GGUF lesliemore · text-to-image Open weights — — ✓ apache-2.0 108 K from 8.4 GB
1039 opus-mt-ar-en Helsinki-NLP · translation Open weights — 512 ✓ apache-2.0 107 K —
1040 Krea-2-Raw krea · text-to-image · gated Open weights 12.8 B — other 107 K from 70.6 GB
1041 janku-v5-nsfw-trained-noobai-rou-wei-illustrious-xl-v50-sdxl John6666 · text-to-image Open weights 2.6 B — other 107 K from 8.5 GB
1042 Mistral-Small-3.2-24B-Instruct-2506 mistralai · text-generation Open weights Also via API 24.0 B — ✓ apache-2.0 105 K from 109.8 GB
1043 opus-mt-ko-en Helsinki-NLP · translation Open weights — 512 ✓ apache-2.0 103 K —
1044 HunyuanDiT-v1.1-Diffusers-Distilled Tencent-Hunyuan · text-to-image Open weights 1.5 B — other 101 K from 16.7 GB
1045 HiDream Comfy-Org · text-to-image Open weights — — ✓ mit 100 K from 262.9 GB
1046 Qwen-Image-2.1-Turbo-Quantized abenzerps · text-to-image Open weights — — other 100 K from 5.0 GB
1047 Qwen2.5-VL-72B-Instruct Qwen · image-text-to-text Open weights Also via API 73.4 B 128 K other 99 K from 173.0 GB
1048 dreamshaper-xl-v2-turbo Lykon · text-to-image Open weights 2.6 B — ⚠ openrail++ 99 K from 46.7 GB
1049 RealVisXL_V5.0_Lightning SG161222 · text-to-image Open weights 2.6 B — ⚠ openrail++ 98 K from 46.7 GB
1050 FLUX.1-schnell-GGUF unsloth · text-to-image Open weights 11.9 B — ✓ apache-2.0 98 K from 6.7 GB
VRAM figures are estimates for the smallest available quantization at 8K context — see /methodology. Downloads refresh nightly from the Hugging Face API; API prices (blended 3:1 input:output per 1M tokens) and Artificial Analysis indices from OpenRouter.