Skip to content

EthenEthenEthen

LoginPlatformTry Ethen

Open Source Model Profile · zai-org

GLM-5.3-Flash-BF16

GLM-5.3-Flash-BF16 is a 321.32B-parameter multimodal model from zai-org. Captured evidence documents GLM5 architecture, MIT licensing, and image-text support.

Publisher
zai-org
Task
image-text-to-text
Model type
glm5_next
License
mit
Library
transformers
Publication status
Accepted · not indexed

Model overview

GLM-5.3-Flash-BF16 is published by zai-org as an image-text-to-text model. The captured configuration identifies Glm5NextForConditionalGeneration with about 321.32B parameters under MIT. According to the model card, it is the efficiency-focused Flash entry in the GLM-5 series.

Recorded capabilities

Publisher-described multimodal Flash

According to the model card, GLM-5.3-Flash is the first natively multimodal model in the GLM-5 series.

321.32B BF16 configuration

The captured configuration identifies Glm5NextForConditionalGeneration with 321,323,031,390 parameters and BF16 tensor references.

Publisher-described efficiency design

The card describes hybrid sparse and linear attention with Manifold-Constrained Hyper-Connections and a 30T-token multimodal corpus.

Paper and multilingual tags

Tags reference arXiv 2602.15763 with English and Chinese conversational support.

Use cases in the source record

  • Multimodal image-text workflows for the publisher-described coding and agentic tasks, subject to independent evaluation.
  • Transformers and endpoints-compatible experiments using English and Chinese conversational inputs.

Limitations and unknowns

  • Performance and price comparisons in the card are publisher-reported and have not been independently verified by Ethen.
  • No context-window value was extracted; evaluation footnotes mentioning 300,000 tokens describe test settings, not a verified model limit.
  • Provider state is historical snapshot data, not independently refreshed current availability.

Source and provenance

Source: zai-org/GLM-5.3-Flash-BF16

Captured: Unknown. Processed: 2026-09-07T19:35:01.759307+00:00.

GLM-5.3-Flash-BF16 👋 Join our WeChat or Discord community. 📖 Check out the GLM-5.3-Flash blog and GLM-5 Technical report . 📍 Use GLM-5.3-Flash API services on Z.ai API Platform. Introduction We introduce GLM-5.3-Flash, the first natively multimodal model in the GLM-5 series. With 320B total parameters and just 18B active parameters, it outperforms GLM-5.2 across benchmarks and real-world workloads at one-tenth the price, while approaching Claude Opus 4.8 on coding and agentic benchmarks. GLM-5.3-Flash starts from a newly trained base model, with its architecture and training recipe redesigned around capability and efficiency. For…

F001F002F003F004F005F006F007F008F009F010F012F014F015F016F017