← all releases
kai-os·Apr 28, 2026·V2·models
Carnice-V2-27B
Carnice-V2-27B is a 27 billion parameter open-source AI model built on Qwen3-27B that beats models 10x its size on the Hermes-agent benchmark. Designed to run on consumer GPUs (RTX 3090+), it is trained on more and higher-quality data than its predecessor, available in full and GGUF formats.
open-sourceLLM27BQwen3consumer-GPUagent-modelGGUF
overview
Carnice-V2-27B is a 27 billion parameter open-source AI model built on Qwen3-27B that beats models 10x its size on the Hermes-agent benchmark. Designed to run on consumer GPUs (RTX 3090+), it is trained on more and higher-quality data than its predecessor, available in full and GGUF formats.
key features
- 0127B parameters beating models 10x larger on Hermes-agent benchmark
- 02Runs on consumer GPUs — RTX 3090 and above
- 03Available in full precision and GGUF quantized formats
related productsbrowse all →
Protenixopen sourceProtenix is an open-source, trainable PyTorch implementation of AlphaFold 3, designed for high-accuracy biomolecular structure prediction. It aims to advance accessible and extensible research tools for the computational biology community. ([github.com](https://github.com/bytedance/Protenix?utm_source=openai))Z-Imageopen sourceZ-Image is an efficient image generation foundation model developed by Tongyi-MAI, designed to produce high-quality, diverse, and stylistically versatile images. It serves as a robust base for creators, researchers, and developers seeking advanced image generation capabilities.