GLM 5.3 FlashX AI Evaluation | LayerLens - Benchmark Results & Performance

Processing your request

This may take a moment. We're preparing the latest results.

Evaluations Compass Benchmarks Spaces

Learning Docs

Processing your request

This may take a moment. We're preparing the latest results.

GLM 5.3 FlashX AI Evaluation | LayerLens - Benchmark Results & Performance

Model

GLM 5.3 FlashX by z-ai

Text+image+video->text

Free credits included. No credit card required.

Run your own evaluation

GLM-5.3-FlashX is the high-speed variant of Z.ai's GLM-5.3-Flash, delivering inference speeds of up to 200 tokens/s. It is the first natively multimodal model in the GLM-5 series, built on a 320B-total/18B-active Mixture-of-Experts architecture with a novel hybrid sparse and linear attention design that reduces attention computation and KV cache costs significantly. The model supports a 1M-token context window with native image and video input, and excels at visual coding, agentic workflows, Office document generation, and long-context tasks. It is released under an MIT license with open weights on Hugging Face.

Evaluations

No evaluations run yet

Latency

—

Context length

tokens

Cost

$0.37 · $1.25

input · output per 1M tokens

Key takeaways

GLM-5.3-FlashX is a high-speed, natively multimodal model designed for quick inference (up to 200 tokens/s) across various tasks including visual coding, agentic workflows, document generation, and long-context scenarios.
It is built on a 320B-total/18B-active Mixture-of-Experts (MoE) architecture with a novel hybrid sparse and linear attention design that optimizes computation and KV cache costs.
The model supports a 1M-token context window and handles native image and video inputs, making it versatile for complex multimodal tasks.
Key capabilities include strong performance in visual coding, agentic workflows, Office document generation, and long-context understanding due to its large context window and multimodal nature.
The model is open-weights on Hugging Face under an MIT license, indicating accessibility for research and development, and is provided by Z.ai with a focus on high-speed performance.

Technical specifications

Context length: 1M tokens
Max output: —
Architecture: MoE
Modalities: text+image+video->text
Parameters: ~320B
Open weights: Yes
Licence: MIT
Regions: China
Vendor: z-ai
Provider model key: z-ai/glm-5.3-flashx