AImodelsComparison.com 1,226 open · 200 API
Menu
Models Best of Trending Can I run it? Methodology About
1,226 models · 1,226 open weights · 200 API-only · refreshed nightly

Open-weights models

Models whose weights are published on Hugging Face — download them and run them on your own hardware. Many are also served by hosted providers; those carry an "Also via API" badge with live prices.

Type
All 1,426 Open weights (self-host) 1,226 API-only 200
Task
All text-generation 534 text-to-image 155 image-text-to-text 112 ASR 86 sentence-similarity 64 translation 56 text-classification 51 feature-extraction 48 image-classification 39 token-classification 21 text-to-speech 19 audio-classification 12 image-segmentation 11 summarization 8 object-detection 8 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
$
534 models
#ModelParamsContextCommercial use30dMin VRAM
01 Qwen3-0.6B Qwen · text-generation Open weights 750 M 41 K ✓ apache-2.0 25.2 M from 2.3 GB
02 gpt2 openai-community · text-generation Open weights 140 M — ✓ mit 18.8 M from 1.1 GB
03 Qwen3-4B Qwen · text-generation Open weights 4.0 B 41 K ✓ apache-2.0 11.4 M from 10.0 GB
04 Qwen3-14B Qwen · text-generation Open weights Also via API 14.8 B 41 K ✓ apache-2.0 9.4 M from 35.2 GB
05 Qwen3-8B Qwen · text-generation Open weights 8.2 B 41 K ✓ apache-2.0 8.1 M from 19.7 GB
06 Qwen2.5-7B-Instruct Qwen · text-generation Open weights Also via API 7.6 B 33 K ✓ apache-2.0 6.4 M from 18.4 GB
07 gpt-oss-20b openai · text-generation Open weights Also via API 21.5 B 131 K ✓ apache-2.0 5.7 M from 34.0 GB
08 opt-125m facebook · text-generation Open weights 125 M 2 K other 5.6 M from 0.8 GB
09 DeepSeek-V3.2 deepseek-ai · text-generation Open weights Also via API 685.4 B 164 K ✓ mit 5.4 M from 861.7 GB
10 Qwen2.5-0.5B-Instruct Qwen · text-generation Open weights 490 M 33 K ✓ apache-2.0 5.3 M from 1.7 GB
11 Ornith-1.5-9B-GGUF ornith-ai · text-generation Open weights — — ✓ mit 5.0 M from 6.9 GB
12 Ornith-1.0-9B-GGUF deepreinforce-ai · text-generation Open weights — — ✓ mit 4.9 M from 6.7 GB
13 Llama-3.1-8B-Instruct meta-llama · text-generation · gated Open weights Also via API 8.0 B — ⚠ llama3.1 4.9 M from 19.4 GB
14 Qwen2.5-1.5B-Instruct Qwen · text-generation Open weights 1.5 B 33 K ✓ apache-2.0 4.8 M from 4.1 GB
15 dolphin-2.9.1-yi-1.5-34b dphn · text-generation Open weights 34.4 B 8 K ✓ apache-2.0 4.5 M from 81.3 GB
16 DeepSeek-V4-Flash-0731 deepseek-ai · text-generation Open weights Also via API 304.2 B 1.0 M ✓ mit 4.3 M from 229.7 GB
17 Qwen3-1.7B Qwen · text-generation Open weights 2.0 B 41 K ✓ apache-2.0 4.1 M from 5.3 GB
18 Qwen3.6-35B-A3B-NVFP4 nvidia · text-generation Open weights 18.7 B — ✓ apache-2.0 4.1 M from 29.1 GB
19 Llama-3.2-1B-Instruct meta-llama · text-generation · gated Open weights Also via API 1.2 B — ⚠ llama3.2 4.0 M from 3.4 GB
20 Ornith-1.5-35B-A3B-GGUF ornith-ai · text-generation Open weights — — ✓ mit 3.8 M from 24.4 GB
21 Qwen2.5-3B-Instruct Qwen · text-generation Open weights 3.1 B 33 K other 3.8 M from 7.8 GB
22 Ornith-1.0-35B-GGUF deepreinforce-ai · text-generation Open weights — — ✓ mit 3.8 M from 23.8 GB
23 Kimi-K3-DSpark RadixArk · text-generation Open weights 2.3 B 1.0 M unknown 3.7 M from 5.8 GB
24 pythia-160m EleutherAI · text-generation Open weights 210 M 2 K ✓ apache-2.0 3.3 M from 0.9 GB
25 gpt-oss-120b openai · text-generation Open weights Also via API 120.4 B 131 K ✓ apache-2.0 2.9 M from 162.1 GB
26 Ornith-1.0-35B deepreinforce-ai · text-generation Open weights <0.1 M — ✓ mit 2.8 M from 77.7 GB
27 Qwen3-32B Qwen · text-generation Open weights Also via API 32.8 B 41 K ✓ apache-2.0 2.8 M from 77.5 GB
28 Qwen3-4B-Instruct-2507 Qwen · text-generation Open weights 4.0 B 262 K ✓ apache-2.0 2.6 M from 10.0 GB
29 Qwen3-14B-AWQ Qwen · text-generation Open weights 14.8 B 41 K ✓ apache-2.0 2.6 M from 13.7 GB
30 Qwen3.8-27B-Uncensored-GGUF JonathanColetti · text-generation Open weights — — ✓ apache-2.0 2.4 M from 1.5 GB
31 Ornith-1.0-9B deepreinforce-ai · text-generation Open weights <0.1 M — ✓ mit 2.4 M from 21.2 GB
32 Ornith-1.0-9B-GGUF ornith-ai · text-generation Open weights — — ✓ mit 2.4 M from 6.7 GB
33 Ternary-Bonsai-2-27B-gguf prism-ml · text-generation Open weights Also via API 27.0 B — ✓ apache-2.0 2.3 M from 5.2 GB
34 JiRackUltra_1b CMSManhattan · text-generation Open weights 1.8 B 131 K ✓ mit 2.3 M from 1.8 GB
35 NVIDIA-Nemotron-3-Nano-4B-BF16 nvidia · text-generation Open weights 4.0 B 262 K other 2.1 M from 9.8 GB
36 Qwen3-8B-AWQ Qwen · text-generation Open weights 8.2 B 41 K ✓ apache-2.0 2.1 M from 8.4 GB
37 Qwen2.5-7B-Instruct-AWQ Qwen · text-generation Open weights 7.6 B 33 K ✓ apache-2.0 2.0 M from 7.8 GB
38 GLM-5.3 zai-org · text-generation Open weights Also via API 753.3 B 1.0 M other 1.9 M from 944.7 GB
39 SmolLM2-135M HuggingFaceTB · text-generation Open weights 130 M 8 K ✓ apache-2.0 1.8 M from 0.8 GB
40 Mistral-7B-Instruct-v0.2 mistralai · text-generation Open weights 7.2 B 33 K ✓ apache-2.0 1.7 M from 17.5 GB
41 distilgpt2 distilbert · text-generation Open weights 90 M — ✓ apache-2.0 1.7 M from 0.9 GB
42 Qwen2.5-Coder-14B-Instruct Qwen · text-generation Open weights 14.8 B 33 K ✓ apache-2.0 1.7 M from 35.2 GB
43 Qwen2.5-VL-7B-Instruct-NVFP4 nvidia · text-generation Open weights 5.0 B 128 K other 1.6 M from 9.2 GB
44 Ornith-1.0-35B-GGUF ornith-ai · text-generation Open weights — — ✓ mit 1.5 M from 23.8 GB
45 Qwen2.5-Coder-7B-Instruct Qwen · text-generation Open weights 7.6 B 33 K ✓ apache-2.0 1.5 M from 18.4 GB
46 DeepSeek-R1 deepseek-ai · text-generation Open weights 684.5 B 164 K ✓ mit 1.4 M from 860.6 GB
47 Ornith-1.0-35B ornith-ai · text-generation Open weights <0.1 M — ✓ mit 1.4 M from 77.7 GB
48 pythia-70m-deduped EleutherAI · text-generation Open weights 100 M 2 K ✓ apache-2.0 1.4 M from 0.7 GB
49 Swift-1.5-Qwen3.8-27B-GSQ-RCO-GGUF ukisai · text-generation Open weights 27.0 B — other 1.4 M from 4.6 GB
50 Qwen2.5-14B-Instruct-AWQ Qwen · text-generation Open weights 14.8 B 33 K ✓ apache-2.0 1.4 M from 13.7 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.