Z.ai (Zhipu) logo
Z.ai (Zhipu) · China · Released 2026-10

GLM-5.3

Low-cost coding specialist with strong anti-contamination benchmarks

Price in / 1M$0.60
Price out / 1M$2.20
Context window200K
Max output32768
LicenseOpen weights

Overview

GLM-5.3 improves roughly 50% over its predecessor on Z.ai’s internal code bench and scored highly on SWE-bench Pro, a contamination-resistant engineering benchmark. With open weights and low API pricing, it is a compelling option for budget coding agents and domestic Chinese deployment.

Best for

  • Coding agents
  • Budget automation
  • Chinese enterprise

Strengths & limitations

✓ Strengths

  • Strong SWE-bench Pro
  • Open weights
  • Low price

△ Limitations

  • Shorter context than flagships
  • Smaller multimodal footprint

Reported benchmarks

BenchmarkScoreNote
SWE-bench Pro~62—
DeepSWE66.9—

Benchmark figures are reported values from public sources and may use different evaluation harnesses. See sources below.

Frequently asked questions

How much does GLM-5.3 cost?

The API price is $0.60 per 1M input tokens and $2.20 per 1M output tokens (as of 2026-11-02). Prices change — check the official provider page for current rates.

Is GLM-5.3 open source?

GLM-5.3 is open weights. Weights are available, subject to the licence terms — review the model card before commercial use.

What is the context window of GLM-5.3?

GLM-5.3 supports a context window of 200K tokens with a maximum output of 32768 tokens per response.

What is GLM-5.3 best at?

Based on our scoring, GLM-5.3 is strongest at strong swe-bench pro, open weights, low price. It is a common choice for coding agents, budget automation, chinese enterprise.

Sources

Similar models