utDeepSeek · Downloadable
DeepSeek-V4-Flash-0731
Generates text for conversations and writing.
- Source listed
- Jul 31, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- mit
WHAT’S NEW
New AI models.
Keep up without the noise.
The latest source listings from makers we track, covering language, images, video, audio, and 3D. Catalog checked Sep 10, 2026.
Sorted by the date a source repository was created, which may precede public release. This is a feed of new listings, not confirmed release announcements. Downloadable weights do not automatically mean an open-source license. Coverage and date meanings.
Latest listings
1,388 dated recordsNVIDIA · Downloadable
NVIDIA-NemotronLabs-VoiceChat-11B
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jul 29, 2026
- Parameters
- 11B
- 4-bit weights only
- ≈ 5.5 GB
- License
- openmdw-1.1
MiniMax · Downloadable
MiniMax-H3
Generates video. Resolution, clip length, and speed depend on the version and setup.
- Source listed
- Jul 28, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- other
Lightricks · Downloadable
LTX-2.3-22b-IC-LoRA-Relight
Generates video. Resolution, clip length, and speed depend on the version and setup.
- Source listed
- Jul 27, 2026
- Parameters
- 22B
- 4-bit weights only
- ≈ 11 GB
- License
- other
Lightricks · Downloadable
LTX-2.5-Diffusers
Generates video. Resolution, clip length, and speed depend on the version and setup.
- Source listed
- Jul 26, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- other
Microsoft · Downloadable
Mage-ViT
Turns content into numerical representations for semantic search and retrieval.
- Source listed
- Jul 26, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- mit
Microsoft · Downloadable
Mage-VL
Understands text and images and responds in text.
- Source listed
- Jul 25, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- apache-2.0
Microsoft · Downloadable
VibeVoice-ASR-BitNet
Works with speech or audio. Check the model card for transcription and generation capabilities.
- Source listed
- Jul 24, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- mit
Lightricks · Downloadable
LTX-2.5
Generates video. Resolution, clip length, and speed depend on the version and setup.
- Source listed
- Jul 23, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- other
NVIDIA · Downloadable
DeepSeek-V4-Pro-nvfp4-DSpark
Generates text for conversations and writing.
- Source listed
- Jul 22, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- mit
NVIDIA · Downloadable
DeepSeek-V4-Flash-nvfp4-DSpark
Generates text for conversations and writing.
- Source listed
- Jul 22, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- mit
NVIDIA · Downloadable
Cosmos-H-Dreams
Generates video. Resolution, clip length, and speed depend on the version and setup.
- Source listed
- Jul 21, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- other
Lightricks · Downloadable
LTX-2.3-22b-IC-LoRA-Clean-Plate
Generates video. Resolution, clip length, and speed depend on the version and setup.
- Source listed
- Jul 19, 2026
- Parameters
- 22B
- 4-bit weights only
- ≈ 11 GB
- License
- other
Microsoft · Downloadable
Fara1.5-27B
Understands text and images and responds in text.
- Source listed
- Jul 17, 2026
- Parameters
- 27B
- 4-bit weights only
- ≈ 13.5 GB
- License
- mit
Microsoft · Downloadable
Fara1.5-4B
Understands text and images and responds in text.
- Source listed
- Jul 17, 2026
- Parameters
- 4B
- 4-bit weights only
- ≈ 2 GB
- License
- mit
NVIDIA · Downloadable
NV-JEPA-DNA-HyenaDNA
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jul 16, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- other
NVIDIA · Downloadable
NV-JEPA-DNA-NTv3
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jul 16, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- other
NVIDIA · Downloadable
NV-JEPA-DNA-DNABERT2
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jul 16, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- other
Mistral AI · Downloadable
Shieldstral-1.0-3B
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jul 16, 2026
- Parameters
- 3B
- 4-bit weights only
- ≈ 1.5 GB
- License
- apache-2.0
NVIDIA · Downloadable
MiniMax-M3-DSpark
Generates text for conversations and writing.
- Source listed
- Jul 15, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- other
NVIDIA · Downloadable
Ising-Decoder-ColorCode-1-Fast
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jul 15, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- openmdw-1.1
Lightricks · Downloadable
LTX-2.3-22b-LoRA-Foley-V2A
Works with speech or audio. Check the model card for transcription and generation capabilities.
- Source listed
- Jul 15, 2026
- Parameters
- 22B
- 4-bit weights only
- ≈ 11 GB
- License
- other
Microsoft · Downloadable
bitnet-embedding-0.6b
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jul 15, 2026
- Parameters
- 600M
- 4-bit weights only
- ≈ 0.3 GB
- License
- mit
Microsoft · Downloadable
bitnet-embedding-270m
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jul 15, 2026
- Parameters
- 270M
- 4-bit weights only
- ≈ 0.1 GB
- License
- mit
What makes this feed useful?
Capability tags distinguish a model that reads images from one that creates them. Each listing links to its source and a permanent profile, so you can inspect its license and compare it with a model you already know.
Does this cover every new release?
No. The current feed tracks public Hugging Face metadata from selected makers. It can miss untracked makers, cloud-only releases, and records beyond a source’s retrieval limit. Announced models without a source listing are not included. Check Sources & methodology for the current coverage.
Listings update with the catalog automatically. A missing or delayed source can leave older information visible; each model profile shows when it was checked.