Hugging FacePublisher: beyoru

Kiwen1.1-27B

Model overview

What this model is for

Kiwen1.1-27B is a 27-billion-parameter image-text-to-text model, listed as a fine-tune of Qwen3.8-27B, and is mainly useful for coding, reasoning, vision-enabled workflows, and agentic applications. Its recorded capabilities include Transformers and safetensors support, conversational interaction, Vietnamese, mathematics, coding, reasoning, and agentic use. The listed modality indicates that it can accept image-and-text inputs and produce text outputs, making it relevant to multimodal assistants, visual question answering, coding support, vibe-coding, and task-oriented agents. The model record also tags it for Kimi-related work and distillation, although the available source content does not provide implementation details or evaluation results for those uses. The repository is associated with the Apache 2.0 license. Context length, inference speed, and input/output pricing are not disclosed. The available Hugging Face page content does not expose benchmark scores, safety characteristics, deployment requirements, or a dated model-card update, so performance and operational suitability should be validated independently before production use. The Hugging Face repository is the authoritative source reference provided for this listing.

Classification

Categories, modalities and capabilities

Input and output modalities

imagetext

Supported capabilities

agentic

Continue exploring