GLM-5.3 Flash

by Z.ai

GLM-5.3-Flash is Z.AI's first natively multimodal model in the GLM-5 series. It has 320B total parameters, with 18B active, and a hybrid attention design intended to lower the cost of long prompts.

The model can work with images and video, use browser and computer tools, and inspect rendered interfaces. Its 1,048,576-token context window suits coding agents, visual development, and other tasks that need a lot of context. It is cheaper to run than GLM-5.2, while benchmark results show stronger performance across the tests Z.AI reports.

Best for: Lower-cost coding and visual tasks with a million-token context.

1M context

Text
Reasoning
Tool use
Image input
Vision
Video input

Model weights

Try it in the playground

Pricing

USD per 1M tokens. Cheapest input plus output first (refreshed hourly).

Provider	Input	Cached input	Output
Boundless	$0.12	$0.024	$0.40
Baseten	$0.15	$0.03	$0.50
DigitalOcean	$0.15	$0.03	$0.50
Fireworks	$0.15	$0.03	$0.50
Together AI	$0.15	$0.03	$0.50

What one request costs

Prompt94Ktokens

Cached prefix66Ktokens

Reply2Ktokens

$0.0056

per request to GLM-5.3 Flash

Prompt: 60.8%
Cached: 28.4%
Reply: 10.8%

The same request, across other models