AI models
New open AI models
New on Hugging Face: what each one does and the smallest computer that runs it.
Today 1 new model
-
Makes imageslodestones
Kromoss
A sharper fine-tune of MOSS-Image-Base that makes 1024 px images and fixes the original model's blur.
- A drop-in replacement for MOSS-Image-SFT
- Ships a fast add-on that makes an image in 8 steps instead of 50
- Works in ComfyUI out of the box
Minimum 8GB graphics card, or 16GB computer memory (slower) e.g. GeForce RTX 5050
Free for commercial use
Download
Sunday, October 18 1 new model
-
Writes down speechAuroraAI-Research
AuroraASR-1M
A tiny model that turns English speech into text, with under one million parameters.
- On a single CPU core it runs about 436 times faster than real time
- About 1 MB in its smallest version
- English only, and less accurate than Whisper Tiny
Minimum A recent phone or any computer
Free for commercial use
Download
Saturday, October 17 1 new model
-
Reads imagesdeepseek-ai
DeepSeek-V4.1-Flash-1017
An update to DeepSeek-V4.1-Flash with better reasoning, coding, agent tasks and image understanding.
- Terminal-Bench 4.0 score up from 31.2 to 42.3
- Understands text and images together
- You can try it in DeepSeek's chat without downloading anything
Minimum Server with several graphics cards
Free for commercial use
Download
Friday, October 16 1 new model
-
Chat and writingnvidia
NVIDIA-Nemotron-Labs-Dory1.0-4B-BF16
A small NVIDIA language model for general reasoning, built to run on modest hardware.
- Reads long texts, up to 256K tokens
- Runs part of its layers several times to think deeper at a small size
- English only
Minimum 8GB graphics card, or 16GB computer memory (slower) e.g. GeForce RTX 5050
Free for commercial use
Download
Thursday, October 15 4 new models
-
Chat and writingtencent
Hy-Memory-30B-A3B
A Tencent model that writes long-term memory for AI agents by picking what to keep from their conversations.
- Pulls out the facts worth remembering from conversations and words them clearly
- 30B parameters in total, only about 3B active per token
- Built to plug into memory tools like Mem0 and MemOS, not for general chat
Minimum 24GB graphics card e.g. GeForce RTX 5090or a computer with 32GB memory and no graphics card (slower)
Free for commercial use
Download -
Speaks textSupraLabs
SupraTTS-1.0-Preview
A very small model that turns English text into speech, with about one million parameters.
- On a plain CPU it runs about 90 times faster than real time, the maker says
- Needs only Python and three pip packages to run
- English only, a preview release, and the quality scores are the maker's own
Minimum A recent phone or any computer
Free for commercial use
Download -
3D shapesAlayaLab
MeshAxiom-Base
A model that turns one picture of an object into a 3D mesh that can be saved as OBJ or GLB.
- Input is a 512×512 RGB image on a black background
- Exports OBJ, JSON and GLB; the requested face count is only a guide
- Weights only: it needs the separate MeshAxiom code package to run
Minimum 24GB graphics card e.g. GeForce RTX 5090
Check the licence before use
Download
The minimum is our estimate from the model’s size: a 4-bit copy for text models, 8-bit for image, video and sound. Licences and sizes come from each maker’s page; read the licence before using a model in your work.
