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 29, 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

3,178 dated records

DeepSeek · Downloadable

DeepSeek-V4-Flash-Base

Text generation

Generates text, according to its published model architecture (DeepseekV4ForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Apr 22, 2026
Parameters
292B
4-bit weights only
≈ 146 GB
License
Not disclosed
Compare this model →

Google · Cloud

Gemini Embedding 2

VisionVideo understandingAudio understanding

Cloud API for vision, video understanding, audio understanding, chat & writing. Capabilities are catalog-listed, not independently scored.

Source listed
Apr 22, 2026
License
Provider terms
Compare this model →

OpenMOSS · Downloadable

MOSS-Music-8B-Instruct

Capabilities unconfirmed

Works with speech or audio. Check the model card for transcription and generation capabilities.

Source listed
Apr 21, 2026
Parameters
8B
4-bit weights only
≈ 4 GB
License
apache-2.0
Compare this model →

OpenMOSS · Downloadable

MOSS-Music-8B-Thinking

Capabilities unconfirmed

Works with speech or audio. Check the model card for transcription and generation capabilities.

Source listed
Apr 21, 2026
Parameters
8B
4-bit weights only
≈ 4 GB
License
apache-2.0
Compare this model →

inclusionAI · Downloadable

DR-Venus-4B-RL

Text generation

Generates text, according to its published model architecture (Qwen3ForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Apr 21, 2026
Parameters
4B
4-bit weights only
≈ 2 GB
License
Not disclosed
Compare this model →

Qwen · Alibaba · Downloadable

Qwen3.6-27B

Vision

Understands text and images and responds in text.

Source listed
Apr 21, 2026
Parameters
27B
4-bit weights only
≈ 13.5 GB
License
apache-2.0
Compare this model →

inclusionAI · Downloadable

DR-Venus-4B-SFT

Text generation

Generates text, according to its published model architecture (Qwen3ForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Apr 21, 2026
Parameters
4B
4-bit weights only
≈ 2 GB
License
Not disclosed
Compare this model →

NVIDIA · Downloadable

Kumo-Anomaly

Capabilities unconfirmed

A specialized research model. See the model card for its intended task and supported setup.

Source listed
Apr 21, 2026
Parameters
Not reported
4-bit weights only
Not reported
License
apache-2.0
Compare this model →

NVIDIA · Downloadable

nv-tesseract-ad-diffusion

Capabilities unconfirmed

A specialized research model. See the model card for its intended task and supported setup.

Source listed
Apr 21, 2026
Parameters
Not reported
4-bit weights only
Not reported
License
apache-2.0
Compare this model →

NVIDIA · Downloadable

omni-dreams-models

Video generation

Generates video. Resolution, clip length, and speed depend on the version and setup.

Source listed
Apr 21, 2026
Parameters
Not reported
4-bit weights only
Not reported
License
other
Compare this model →

Google · Preview

Deep Research Max Preview (Apr-21-2026)

VisionVideo understandingImage generationImage editingAudio understandingReasoningTool use

Cloud API for vision, video understanding, image generation, image editing, audio understanding, chat & writing, reasoning, tool use. Capabilities are catalog-listed, not independently scored.

Source listed
Apr 21, 2026
License
Provider terms
Compare this model →

Google · Preview

Deep Research Preview (Apr-21-2026)

VisionVideo understandingImage generationImage editingAudio understandingReasoningTool use

Cloud API for vision, video understanding, image generation, image editing, audio understanding, chat & writing, reasoning, tool use. Capabilities are catalog-listed, not independently scored.

Source listed
Apr 21, 2026
License
Provider terms
Compare this model →

OpenAI · Cloud

gpt-image-2

Image generationImage editing

Cloud API for image generation, image editing. Capabilities are catalog-listed, not independently scored.

Source listed
Apr 21, 2026
License
Provider terms
Compare this model →

Moondream · Downloadable

tracking

Capabilities unconfirmed

A specialized research model. See the model card for its intended task and supported setup.

Source listed
Apr 19, 2026
Parameters
194.2M
4-bit weights only
≈ 0.1 GB
License
Not disclosed
Compare this model →

OpenAI · Downloadable

privacy-filter

Token classification

Hugging Face lists its task as token classification. See the model card for supported inputs and setup.

Source listed
Apr 17, 2026
Parameters
1.4B
4-bit weights only
≈ 0.7 GB
License
apache-2.0
Compare this model →

xAI · Cloud

Grok 4.3

VisionReasoningTool use

Cloud API for vision, chat & writing, reasoning, tool use. Capabilities are catalog-listed, not independently scored.

Source listed
Apr 17, 2026
License
Provider terms
Compare this model →

xAI · Cloud

Grok Build 0.1

VisionReasoningTool use

Cloud API for vision, chat & writing, reasoning, tool use. Capabilities are catalog-listed, not independently scored.

Source listed
Apr 16, 2026
License
Provider terms
Compare this model →

Kyutai · Downloadable

moshi-rag-artifacts

Capabilities unconfirmed

A specialized research model. See the model card for its intended task and supported setup.

Source listed
Apr 15, 2026
Parameters
Not reported
4-bit weights only
Not reported
License
Not disclosed
Compare this model →

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.