Jackrong has published Qwopus3.8-27B-Flash-V2-GGUF on Hugging Face. The repository describes it as a renewed post-training release of Qwopus3.8-27B-Flash, itself built on Qwen3.8-27B. It is licensed under Apache-2.0 and is tagged for GGUF and Safetensors distribution.
The model card positions V2 for coding, tool use, function calling, multimodal image-text work, and longer agent loops. It says the update uses revised reward functions and reinforcement learning, with improved Python formatting. The author recommends xhigh reasoning and temperature 0.3 for coding. Usage examples are provided for Transformers, vLLM, SGLang, and Docker Model Runner.
The stated goal is to retain capability while reducing ineffective reasoning and making repeated agent calls more practical. The repository does not establish independently verified benchmark gains, latency, inference cost, hardware requirements, or performance against the underlying Qwen3.8-27B model. Its reported results and workload claims are those of the model card author.