GLM-5.3
Z.ai · #1 in our ranking · Best overall
GLM-5.3 is an open-weight mixture-of-experts model from Z.ai, released Aug 2026, with 753B parameters, a 1M-token context window, and the GLM-5.3 License.
GLM-5.3 keeps the GLM-5.2 base model and puts every gain into post-training. Z.ai reports open-source state of the art on Terminal Bench 3.0 and Agents' Last Exam, and a 50% gain over GLM-5.2 on its in-house code benchmark.
GLM-5.3 specs
- Provider
- Z.ai
- Architecture
- Mixture-of-experts
- Parameters
- 753B total
- Context
- 1M tokens | 786K words
- License
- GLM-5.3 License
- Input
- Text
- Released
- Aug 2026
- Weights
- Hugging Face
Sourced from the model card on Hugging Face. Last checked 2026-10-01.
Highlights
- 1M-token context
- Top open-weight scores on agentic benchmarks
- Built for long, multi-step tool use
Where it fits for payment teams
- Payment reconciliation
#2 pick. Top open-weight agentic scores, so it reliably chains ledger queries, processor APIs, and file parsing across a long run.
- Chargebacks and disputes
#3 pick. Strong at long, multi-tool runs that pull from the processor, shipping data, and the help desk.
- Merchant underwriting
#1 pick. The strongest open-weight model for long agentic runs that chain KYB, sanctions, and internal data.
- AML and compliance investigations
#2 pick. Reliable over long investigations that touch many data sources.
Harnesses in Runtime
Run GLM-5.3 through any of these harnesses on Runtime.
Provider-reported benchmarks
| Terminal Bench 2.1 | 88.2 |
|---|---|
| Toolathlon Verified | 73.0 |
| CyberGym | 84.5 |
Benchmark scores are reported by each model provider on its Hugging Face model card. They are not independently verified by Runtime.
How to run GLM-5.3
Download the weights
Pull zai-org/GLM-5.3 from Hugging Face. Check the GLM-5.3 License before you deploy.
Serve it in your cloud
Run it with vLLM or SGLang on your own GPUs, or use a managed provider that hosts the model.
Put it to work in Runtime
Point an agent or a single skill at the model through OpenCode, then compare it to your current model with evals.
GLM-5.3 FAQ
What is the context window of GLM-5.3?
+
GLM-5.3 supports a 1M-token context window (1,048,576 tokens), roughly 786K words of English text.
How many parameters does GLM-5.3 have?
+
GLM-5.3 is a mixture-of-experts model with 753B total parameters.
Is GLM-5.3 free for commercial use?
+
GLM-5.3 is released under the GLM-5.3 License, a custom license. Commercial use may come with conditions, so read the license on Hugging Face before you deploy or fine-tune it.
What can GLM-5.3 take as input?
+
GLM-5.3 accepts text input and generates text.
Is GLM-5.3 good for payment reconciliation?
+
Yes. GLM-5.3 is our #2 pick for payment reconciliation. Top open-weight agentic scores, so it reliably chains ledger queries, processor APIs, and file parsing across a long run.
Is GLM-5.3 good for chargebacks and disputes?
+
Yes. GLM-5.3 is our #3 pick for chargebacks and disputes. Strong at long, multi-tool runs that pull from the processor, shipping data, and the help desk.
Is GLM-5.3 good for merchant underwriting?
+
Yes. GLM-5.3 is our #1 pick for merchant underwriting. The strongest open-weight model for long agentic runs that chain KYB, sanctions, and internal data.
Is GLM-5.3 good for AML and compliance investigations?
+
Yes. GLM-5.3 is our #2 pick for AML and compliance investigations. Reliable over long investigations that touch many data sources.
Can I run GLM-5.3 in my own cloud with Runtime?
+
Yes. Download the weights from Hugging Face (zai-org/GLM-5.3) and serve them with vLLM or SGLang in your cloud, or use a managed provider that hosts the model. Runtime agents can then use GLM-5.3 through a harness like OpenCode, with your data staying in your environment.