llama.cpp v0.6.0 Released with llama_batch_ext and Metal Boosts
llama.cpp v0.6.0 is out, featuring the new llama_batch_ext API, major Metal backend speedups, web UI updates, and new ...
ggml v0.26.0 Released with Sparse Flash Attention & New Backends
ggml v0.26.0 is released, featuring sparse Flash Attention optimizations, expanded quantization support for WebGPU, a ...
GLM-5.3-Flash-GGUF Vision-Language Model: ~150GB Memory
Discover ggml-org/GLM-5.3-Flash-GGUF, a GGUF quantization of zai-org’s native multimodal MoE model, featuring v ...
OpenJev-GGUF 27.4B Decision-Making Specialized Model: 24GB+ VRAM
Explore OpenJev-GGUF, a GGUF quantized open-weight decision-making model by ggml-org designed for structured choices ...
Magnitude: Up to 2x Faster Than llama.cpp Officially, 21x Slower on Our CPU
Magnitude is an open-source inference engine for AI agents that auto-optimizes kernels for hardware, beating llama.cp ...
Open-Weight Models Weekly: The MiMo-V2.6 Family & Transformers GGUF Support
Weekly roundup of open-weight generative models. Highlights include Qwen3.8-27B, native GGUF support in Transformers, ...
KoboldCpp v1.122 Released with Built-in Agentic Framework
KoboldCpp v1.122 is released, featuring the new KoboldCpp Agent, ubatchsize controls, and upstream fixes from llama.c ...
RamaLama v0.25.0 Released: Security and Engine Updates
RamaLama v0.25.0 is out, featuring secure default bindings, llama.cpp v0.4.1 update, Vulkan support for NVIDIA, and W ...
LFM2.5-VL-3B-DSpark Draft Model for Vision-Language Models: 4GB+ VRAM
LiquidAI has released LFM2.5-VL-DSpark, an experimental DSpark draft model for LFM2.5-VL-3B, speeding up inference wi ...
llama.cpp v0.5.0 Released with Backend and Server Upgrades
llama.cpp v0.5.0 is released, featuring CUDA/Metal hardware acceleration, multi-address server support, and new model ...