GLM-5.3 Flash Uncensored is live

Silk ComputeLaunch1 min read

GLM-5.3 Flash on a dark plate at the center of a burst of white dots, like a camera flash

GLM-5.3 Flash Uncensored is live on Silk Compute, with our lowest input price at $0.24 per 1M tokens. It is Z.ai's fast model with its refusal behavior removed, priced for work you run in volume.

What is it good at?

GLM-5.3 Flash is the one to hand the boring work. At $0.24 per 1M input tokens, it's priced to leave on for background jobs: labeling, data generation, summaries, extraction. Z.ai lists 320B total parameters with just 18B active at a time.

Should you use it in a coding agent?

It works, but check your cache first. If more than about 80% of your input is cached, as with a coding agent, DeepSeek V4.1 Flash 1017 costs less at current prices. Below that, GLM-5.3 Flash wins. Our coding agent guide has the numbers.

Which reasoning effort should you use?

Effort goes from low to max, and high is the default. For a big batch job, try low first. You can always raise it for the few items that need more thought.

Specs and pricing

GLM-5.3 Flash
Model idglm-5.3-flash-uncensored
Context1,048,576 tokens
Max output131,072 tokens
Input$0.24 per 1M tokens
Cached input$0.06 per 1M tokens
Output$1.00 per 1M tokens
Reasoning effortlow, high, max (default high)

Coming from GLM-5.2?

glm-5.3-flash-uncensored replaces glm-5.2-uncensored, which now returns model_not_found. Change the model id.

How do I get started?

Create an API key and send a request to glm-5.3-flash-uncensored at https://api.silkcompute.com. The model page has a prompt you can paste into your coding agent to set it up for you, plus code for Python, JavaScript, cURL, Claude Code, Codex and OpenCode.

More from the blog