New Models

EleutherAI Releases Pythia 160M Weight Seed 2 Corrected

EleutherAI has released the corrected weight-seed 2 version of the Pythia 160M model for reproducibility research.

Image, Video and Audio

Idol 3D Gaussian Splatting Generation Model: ComfyUI Paths

Comfy-Org has released Comfy-Org/Idol, a ComfyUI-compatible repackaging of the open-weight single-image 3D Gaussian S ...

Engines and Tools

NVIDIA Releases AICR for Validated Kubernetes Dynamo Runtimes

Learn how NVIDIA AI Cluster Runtime (AICR) helps build validated, reproducible runtimes for NVIDIA Dynamo on Kubernet ...

Engines and Tools

KoboldCpp 1.124 Released: YuE2, TTS, and Image Gen Upgrades

KoboldCpp v1.124 is out, bringing audio.cpp integration, YuE2 music generation, numerous TTS models, image generation ...

Engines and Tools

llama.cpp Speculative Decoding Test GGUF Models Released

ggml-org released ggml-org/spec-for-testing-GGUF featuring target and draft GGUF models for speculative decoding inte ...

Engines and Tools

Unsloth v0.1.913-beta Released: Faster Updates and Image Gen

Unsloth v0.1.913-beta is released with faster app updates, improved Qwen-Image generation on 16GB VRAM, and Apple Sil ...

New Models

Mage-VL-Cowork Vision-Language Model: 12GB+ VRAM, GGUF Builds

Microsoft has released Mage-VL-Cowork, a 4B hybrid vision-language agent capable of computer use and code execution w ...

New Models

ZwZ2-27B-GGUF Vision-Language Model: 24GB+ VRAM

inclusionAI has released ZwZ2-27B-GGUF, a GGUF version of the vision-language multimodal model ZwZ2-27B, runnable loc ...

Image, Video and Audio

idol 3D Gaussian Splatting Generation Model: File List

Learn about Idol by speridlabs, an open model that generates 3D Gaussian Splatting data from a single image with supp ...

New Models

ZwZ2-4B-GGUF Vision-Language Model: 4GB+ VRAM

GGUF weights for ZwZ2-4B, an open-weight vision-language model based on Qwen3.5-4B, are now available for local execu ...