LocalClaw/LLM/GLM-5.3-Flash
Open-weight MoE

GLM-5.3-Flash

Catalogue summary: Z.ai MIT-licensed GLM-5 refresh with 320B total / 18B active parameters, native multimodal support, hybrid sparse-linear attention and a 1M-token context. Unsloth Dynamic GGUF makes it technically local, but it remains workstation/server-class hardware.

Repository editorial metadata; verify comparative claims in the linked upstream material.

160 GB catalogue minimum160 GB RAMUD-IQ2_XXSCoding assistant
Choose an app

This model needs a special runtime. Unsupported apps are clearly marked.

Compare models
LM StudioNot available for this model
UnslothNot available for this model
OllamaNot available for this model
Open on Hugging FaceFiles, licence and available downloads
llama.cppNot available for this model
Use with LocalClawOptional workspace after the model is installed
This model needs its official runtimeUnsloth Desktop / GLM-5.3 llama.cpp PR
Parameters
320B (18B active, MoE)
Minimum RAM
160 GB
Model size
120 GB
Quantization
UD-IQ2_XXS

Can GLM-5.3-Flash run locally?

GLM-5.3-Flash has a catalogue minimum of 160 GB RAM with UD-IQ2_XXS. Actual memory use and speed vary by context length, runtime, backend and system headroom.

Treat glm-5.3-flash as a server-grade catalogue target. Use the verified artefact with a compatible multi-GPU or distributed runtime and follow its upstream instructions; this is not a one-click desktop LM Studio recommendation.

Model sourcePublic GGUF repository (verified 2026-09-07): unsloth/GLM-5.3-Flash-GGUFUpstream releasePaper or model cardLicence text
chatcodereasoningvisionbeastagenticlong-contextmultimodal

Deployment path

01
Check RAM fitServer-grade target. Plan for 160 GB class multi-GPU memory.
02
Load the modelUse glm-5.3-flash only with a compatible server-grade or distributed runtime; confirm the exact artefact and upstream instructions first.
03
Control locallyUse LocalClaw to manage models, agents, chat, channels and scheduled OpenClaw work.

Catalogue record

  • Family: glm
  • Parameters: 320B (18B active, MoE)
  • Recommended quantization: UD-IQ2_XXS
  • Catalogue minimum RAM: 160 GB
  • Catalogue model size: 120 GB
  • Tags: chat, code, reasoning, vision, beast, agentic, long-context, multimodal

Practical limits

  • Catalogue RAM is a minimum estimate, not a guarantee for every context length or runtime.
  • Speed and memory use vary by quantization, backend, context length and system headroom.
  • Verify architecture, licence and usage restrictions in the linked upstream material before deployment.

Catalogue tags

  • chat
  • code
  • reasoning
  • vision
  • beast
  • agentic

Capability profile

Repository catalogue ratings used by LocalClaw's editorial rubric. They are not a standardized third-party benchmark.

speed
4
quality
10
coding
10
reasoning
10

Technical notes

Developer
Z.ai
License
MIT
Context window
1,000,000 tokens
Architecture
Native multimodal GLM-5 series sparse MoE with 320B total parameters, about 18B active parameters, hybrid sparse-linear attention and Manifold-Constrained Hyper-Connections.

This model fits these next steps

Hardware fit is based on LocalClaw's RAM tier, model size and quantization metadata. Always leave memory headroom for your OS and runtime.

Very large memoryMac Studio Ultra classCheck model size firstNVIDIA GB10 / server optionsMore practical alternativesCompare smaller models

Related catalogue entries

Linked mechanically by family, shared tags and nearby RAM tier; this is not a quality ranking.

GLM-5.2 (744B MoE) 744B (40B active)GLM 4.5 Air (MoE) 106B (12B active, MoE)GLM 4.7 355B (32B active, MoE)GLM 4.6 (355B MoE) 355B (32B active)GLM-4 (32B) 32BGLM 4.7 Flash 14BGLM-4 (9B) 9BGLM-5.1 754B MoE

Where to go next