Skip to content

EthenEthenEthen

LoginPlatformTry Ethen

Open Source Model Profile · zai-org

GLM-4.6V-FP8

GLM-4.6V-FP8 is a 107.75B-parameter vision-language MoE model from zai-org in the GLM-V family. Its model card documents 128k-token training context and native multimodal function calling.

Publisher
zai-org
Task
image-text-to-text
Model type
glm4v_moe
License
mit
Library
transformers
Publication status
Accepted · not indexed

Model overview

GLM-4.6V-FP8 is published by zai-org as an image-text-to-text model with Glm4vMoeForConditionalGeneration architecture and a glm4v_moe model type. Safetensors metadata reports 107,751,931,136 parameters, or about 107.75B. According to the model card, it belongs to the GLM-V multimodal family, and captured metadata records an MIT license with Transformers compatibility.

Recorded capabilities

128k training context

The model card says GLM-4.6V scales its context window to 128k tokens in training.

Native multimodal function calling

The card describes native vision-driven tool use where images and document pages pass directly as tool inputs.

Documented vLLM and SGLang path

The card documents AutoProcessor chat-template inference with vLLM or SGLang and published decoding parameters.

Use cases in the source record

  • Conversational vision-language workflows that submit text plus images through the documented chat-template pipeline.
  • Vision-driven tool-use experiments using the card's native multimodal function-calling capability.

Limitations and unknowns

  • No independent evaluation results were extracted; performance claims in the card are publisher statements, not Ethen measurements.
  • No local hardware or VRAM requirement was extracted from this record.
  • Provider state is historical snapshot data, not independently refreshed current availability.
  • Family, context, and performance claims come from the publisher model card and have not been independently verified by Ethen.

Source and provenance

Source: zai-org/GLM-4.6V-FP8

Captured: Unknown. Processed: 2026-09-07T19:36:04.375561+00:00.

GLM-4.6V This model is part of the GLM-V family of models, introduced in the paper GLM-4.1V-Thinking and GLM-4.5V: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning . GLM-4.6V Blog : https://z.ai/blog/glm-4.6v Paper : https://huggingface.co/papers/2507.01006 GitHub Repository : https://github.com/zai-org/GLM-V Online Demo : https://chat.z.ai/ API Access : Z.ai Open Platform Desktop Assistant App : https://huggingface.co/spaces/zai-org/GLM-4.5V-Demo-App Introduction GLM-4.6V series model includes two versions: GLM-4.6V (106B), a foundation model designed for cloud and high-performance cluster scenarios, and…

F001F002F003F004F005F006F007F010F011F012F013F014F015F016F017