utBRIA AI · Downloadable
FIBO-edit-vlm
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jan 8, 2026
- Parameters
- 4.4B
- 4-bit weights only
- ≈ 2.2 GB
- License
- Not disclosed
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 Oct 30, 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
3,183 dated recordsGoogle · Downloadable
medgemma-1.5-4b-it
Understands text and images and responds in text.
- Source listed
- Jan 7, 2026
- Parameters
- 4B
- 4-bit weights only
- ≈ 2 GB
- License
- other
Microsoft · Downloadable
paza-Phi-4-multimodal-instruct
Works with speech or audio. Check the model card for transcription and generation capabilities.
- Source listed
- Jan 7, 2026
- Parameters
- 5.6B
- 4-bit weights only
- ≈ 2.8 GB
- License
- mit
Tencent · Downloadable
TCAndon-Router
Generates text for conversations and writing.
- Source listed
- Jan 7, 2026
- Parameters
- 4B
- 4-bit weights only
- ≈ 2 GB
- License
- apache-2.0
Tencent · Downloadable
HY-Video-PRFL
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jan 7, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- Not disclosed
Qwen · Alibaba · Downloadable
Qwen3-VL-Reranker-8B
Hugging Face lists its task as text ranking. See the model card for supported inputs and setup.
- Source listed
- Jan 7, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
Qwen · Alibaba · Downloadable
Qwen3-VL-Reranker-2B
Hugging Face lists its task as text ranking. See the model card for supported inputs and setup.
- Source listed
- Jan 7, 2026
- Parameters
- 2B
- 4-bit weights only
- ≈ 1 GB
- License
- apache-2.0
Qwen · Alibaba · Downloadable
Qwen3-VL-Embedding-2B
Turns content into numerical representations for semantic search and retrieval.
- Source listed
- Jan 7, 2026
- Parameters
- 2B
- 4-bit weights only
- ≈ 1 GB
- License
- apache-2.0
Qwen · Alibaba · Downloadable
Qwen3-VL-Embedding-8B
Turns content into numerical representations for semantic search and retrieval.
- Source listed
- Jan 7, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
Kyutai · Downloadable
pocket-tts-without-voice-cloning
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jan 6, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- cc-by-4.0
Supertone · Downloadable
supertonic-2
Works with speech or audio. Check the model card for transcription and generation capabilities.
- Source listed
- Jan 6, 2026
- Parameters
- Not reported
- 4-bit weights only
- Not reported
- License
- openrail
OpenBMB · Downloadable
NOSA-3B
Generates text for conversations and writing.
- Source listed
- Jan 6, 2026
- Parameters
- 3B
- 4-bit weights only
- ≈ 1.5 GB
- License
- apache-2.0
Liquid AI · Downloadable
LFM2.5-1.2B-Instruct
Generates text for conversations and writing.
- Source listed
- Jan 6, 2026
- Parameters
- 1.2B
- 4-bit weights only
- ≈ 0.6 GB
- License
- other
Liquid AI · Downloadable
LFM2-2.6B-Transcript
Generates text for conversations and writing.
- Source listed
- Jan 5, 2026
- Parameters
- 2.6B
- 4-bit weights only
- ≈ 1.3 GB
- License
- other
Microsoft · Downloadable
FrogBoss-32B-2510
Generates text for conversations and writing.
- Source listed
- Jan 5, 2026
- Parameters
- 32B
- 4-bit weights only
- ≈ 16 GB
- License
- mit
Liquid AI · Downloadable
LFM2.5-VL-1.6B
Understands text and images and responds in text.
- Source listed
- Jan 5, 2026
- Parameters
- 1.6B
- 4-bit weights only
- ≈ 0.8 GB
- License
- other
Liquid AI · Downloadable
LFM2.5-Audio-1.5B-GGUF-LEAP
A specialized research model. See the model card for its intended task and supported setup.
- Source listed
- Jan 5, 2026
- Parameters
- 1.5B
- 4-bit weights only
- ≈ 0.8 GB
- License
- Not disclosed
Liquid AI · Downloadable
LFM2.5-1.2B-Base
Generates text for conversations and writing.
- Source listed
- Jan 5, 2026
- Parameters
- 1.2B
- 4-bit weights only
- ≈ 0.6 GB
- License
- other
Tencent · Downloadable
Youtu-LLM-A7B-Base
Understands text and images and responds in text.
- Source listed
- Jan 5, 2026
- Parameters
- 118.4B
- 4-bit weights only
- ≈ 59.2 GB
- License
- other
Liquid AI · Downloadable
LFM2.5-1.2B-JP
Generates text for conversations and writing.
- Source listed
- Jan 4, 2026
- Parameters
- 1.2B
- 4-bit weights only
- ≈ 0.6 GB
- License
- other
BRIA AI · Downloadable
Fibo-Edit
Creates or edits images. Check the model card for supported inputs and output settings.
- Source listed
- Jan 4, 2026
- Parameters
- 8.3B
- 4-bit weights only
- ≈ 4.1 GB
- License
- other
OpenBMB · Downloadable
NOSA-8B
Generates text for conversations and writing.
- Source listed
- Jan 4, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
Lightricks · Downloadable
LTX-2
Generates video. Resolution, clip length, and speed depend on the version and setup.
- Source listed
- Jan 3, 2026
- Parameters
- 18.9B
- 4-bit weights only
- ≈ 9.4 GB
- License
- other
Moonshot AI · Downloadable
Kimi-K2.5
Understands text and images and responds in text.
- Source listed
- Jan 1, 2026
- Parameters
- 1T
- 4-bit weights only
- ≈ 513.4 GB
- License
- other
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.