Meta AI/Vision-Language
Meta's Muse Glimmer 30B Targets Local Agentic Coding
A 30-billion-parameter multimodal model built to run locally, released under Apache 2.0 with an eye on agentic coding workflows.
New TodayDeepSeek
DeepSeek Releases V4-Pro, an MIT-Licensed MoE Model
The company's newest flagship targets reasoning and coding while keeping a permissive open-source license.
Read the story25K downloads
- Size
- 1.7 TB
- Precision
- INT8
- Architecture
- DeepseekV4ForCausalLM
Qwen · Alibaba/Text / LLM
Qwen releases 2.4T-parameter open MoE with 95B active
Alibaba's Qwen team pushes its largest sparse model yet, activating 95 billion parameters per token from a 2.4-trillion-parameter pool.
Qwen · Alibaba/Vision-Language
Qwen3.8-27B Brings Vision to a Dense Model
Alibaba's latest Qwen release pairs image understanding with text under a permissive Apache-2.0 license.
The feed
Latest releases
OpenMOSS/Vision-Language
OpenMOSS Debuts MOSS-VL for Real-Time Vision Interaction
A new open vision-language model family uses gated cross-attention to enable streaming, low-latency multimodal exchanges.
DeepSeek/Text / LLM
DeepSeek Releases V4-Pro-0813 With Open Weights
The Chinese lab pushes a higher-capability checkpoint of its V4 line to Hugging Face under a permissive MIT license.
LiquidAI/Vision-Language
Liquid AI's LFM2.5-VL-3B targets on-device vision
The 3-billion-parameter vision-language model is tuned for faster multimodal work on edge hardware.
NVIDIA/Text → Speech
NVIDIA Opens Magpie TTS for Multilingual Voice Agents
The open-weights text-to-speech model targets low-latency, deployable voice agents across multiple languages.
NVIDIA/Text → Speech
NVIDIA opens Magpie TTS for multilingual voice agents
The company releases open weights for a low-latency, multilingual text-to-speech model aimed at real-time conversational systems.
Cohere/Vision-Language
Cohere Labs releases compact North Micro Vision model
A small multilingual vision-language model built for instruction following arrives under a research-only license.
Momentum
Trending models
Benchmarks
Open LLM leaderboard
| # | Model | Avg. |
|---|---|---|
| 1 | MaziyarPanahi/calme-3.2-instruct-78b | 52.1 |
| 2 | MaziyarPanahi/calme-3.1-instruct-78b | 51.3 |
| 3 | dfurman/CalmeRys-78B-Orpo-v0.1 | 51.2 |
| 4 | MaziyarPanahi/calme-2.4-rys-78b | 50.8 |
| 5 | huihui-ai/Qwen2.5-72B-Instruct-abliterated | 48.1 |
| 6 | Qwen/Qwen2.5-72B-Instruct Qwen · Alibaba | 48.0 |
By type
Newest in each category
Text → Image →
Lodestones/Text → Image
Kroma: An MIT-Licensed Text-to-Image Model for ComfyUI
Lodestones releases an open image generator built on the Krea 2 lineage, aimed squarely at ComfyUI workflows.
Code →
DeepSeek/Text / LLM
DeepSeek Releases V4-Pro, an MIT-Licensed MoE Model
The company's newest flagship targets reasoning and coding while keeping a permissive open-source license.
Text → Video →
MiniMax/Text → Video
MiniMax Releases H3 Video Model on Hugging Face
The company's new diffusion model handles text-to-video and image-to-video, with support for joint audio-video generation.
Vision-Language →
OpenMOSS/Vision-Language
OpenMOSS Debuts MOSS-VL for Real-Time Vision Interaction
A new open vision-language model family uses gated cross-attention to enable streaming, low-latency multimodal exchanges.
On the horizon
Upcoming launches
Announced
Google's DiffusionGemma bets on parallel text generation
An experimental open-weight model swaps token-by-token decoding for discrete diffusion, generating text in parallel blocks.
Google DeepMind·Text / LLM