GLM-5 Models by Zhipu AI
Last refreshed 2026-08-26. Next refresh: weekly.
Details
Capabilities
About
GLM-5 is Zhipu AI (Z.ai)’s flagship foundation-model family for complex systems engineering and long-horizon agent work. Relative to GLM-4.5, Z.ai scales the original GLM-5 checkpoint to about 744B total parameters with about 40B active, grows the pre-training corpus, and adds DeepSeek Sparse Attention to keep long-context serving cheaper. Later GLM-5.1, GLM-5.2, and GLM-5.3 variants stay in the same family with further post-training for coding and long-horizon agents; weights are published on Hugging Face and ModelScope under the zai-org org.
Decision facts
- Best fit
- codingvision and multimodal workreasoning
- Capability starting point
- GLM-5.2 with 1m context and reasoning, JSON / Tool use, and structured outputs
- Lowest tracked input
- GLM-5.3-Flash · $0.075/1M · Z.ai
- Closest related family
- AutoGLM
Current Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
Use when the workload needs coding, 1m context, and reasoning.
Use when the workload needs coding, 1m context, and reasoning.
Use when the workload needs coding, 1m context, and reasoning.
Use when the workload needs 200k context, reasoning, and JSON / Tool use.
Use when the workload needs 200k context, reasoning, and JSON / Tool use.
Use when the workload needs 200k context, reasoning, and JSON / Tool use.
Use when the workload needs 262k context, 9 parameters, and reasoning.
Use when the workload needs 200k context, reasoning, and JSON / Tool use.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| GLM-5.3-Flash | Use when the workload needs coding, 1m context, and reasoning. | 2026-08 | coding1m contextreasoning | Current |
| GLM-5.3 | Use when the workload needs coding, 1m context, and reasoning. | 2026-08 | coding1m contextreasoning | Current |
| GLM-5.2 | Use when the workload needs coding, 1m context, and reasoning. | 2026-06 | coding1m contextreasoning | Current |
| GLM-5.1 | Use when the workload needs 200k context, reasoning, and JSON / Tool use. | 2026-04 | 200k contextreasoningJSON / Tool use | Current |
| GLM-5V-Turbo | Use when the workload needs 200k context, reasoning, and JSON / Tool use. | 2026-04 | 200k contextreasoningJSON / Tool use | Current |
| GLM-5 Turbo | Use when the workload needs 200k context, reasoning, and JSON / Tool use. | 2026-03 | 200k contextreasoningJSON / Tool use | Current |
| GLM-5 9B | Use when the workload needs 262k context, 9 parameters, and reasoning. | 2026-02 | 262k context9 parametersreasoning | Current |
| GLM-5 | Use when the workload needs 200k context, reasoning, and JSON / Tool use. | 2026-02 | 200k contextreasoningJSON / Tool use | Current |
| Zhipu GLM-5 | Use when the workload needs 203k context. | 2026-02 | 203k context | Current |
Release Timeline
5 release groupsSpecifications(9 models)
| Model | Released | Context | Parameters | Vision | Multimodal | Reasoning | JSON / Tool use | Structured Outputs | Code Exec |
|---|---|---|---|---|---|---|---|---|---|
| GLM-5.3-Flash | 2026-08 | 1m | 320B total, 18B active | Yes | Yes | Yes | Yes | Yes | No |
| GLM-5.3 | 2026-08 | 1m | — | No | No | Yes | Yes | Yes | No |
| GLM-5.2 | 2026-06 | 1m | 753B total, 40B active | No | No | Yes | Yes | Yes | Yes |
| GLM-5.1 | 2026-04 | 200k | 754B total, 40B active | No | No | Yes | Yes | Yes | Yes |
| GLM-5V-Turbo | 2026-04 | 200k | 744B total, 40B active | Yes | Yes | Yes | Yes | Yes | No |
| GLM-5 Turbo | 2026-03 | 200k | 744B total, 40B active | No | No | Yes | Yes | Yes | No |
| GLM-5 9B | 2026-02 | 262k | 9 | No | No | Yes | Yes | No | No |
| GLM-5 | 2026-02 | 200k | 744B total, 40B active | No | No | Yes | Yes | Yes | No |
| Zhipu GLM-5 | 2026-02 | 203k | 744B total, 40B active | No | No | No | No | No | No |
Available From(9 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| GLM-5.3-Flash | Z.ai | $0.075 | $0.25 | Serverless |
| GLM-5.3-Flash | OpenRouter | $0.09 | $0.3 | Serverless |
| GLM-5 | OpenRouter | $0.6 | $2.08 | Serverless |
| GLM-5 | Fireworks AI | $1 | $3.2 | Serverless |
| GLM-5 | Together AI | $1 | $3.2 | Serverless |
| GLM-5 | GCP Vertex AI | $1 | $3.2 | Serverless |
| GLM-5 | Vercel AI Gateway | $1 | $3.2 | Serverless |
| GLM-5 | Novita AI | $1 | $3.2 | Serverless |
| GLM-5.1 | OpenRouter | $1.05 | $3.5 | Serverless |
| GLM-5V-Turbo | OpenRouter | $1.2 | $4 | Serverless |
| GLM-5 Turbo | OpenRouter | $1.2 | $4 | Serverless |
| GLM-5 Turbo | Vercel AI Gateway | $1.2 | $4 | Serverless |
| GLM-5V-Turbo | Vercel AI Gateway | $1.2 | $4 | Serverless |
| GLM-5.1 | Novita AI | $1.38 | $4.4 | Serverless |
| GLM-5.1 | Z.ai | $1.4 | $4.4 | Serverless |
| GLM-5.1 | Fireworks AI | $1.4 | $4.4 | Serverless |
| GLM-5.1 | Vercel AI Gateway | $1.4 | $4.4 | Serverless |
| GLM-5.2 | OpenRouter | $1.4 | $4.4 | Serverless |
| GLM-5.3 | Z.ai | $1.4 | $4.4 | Serverless |
| GLM-5.3 | OpenRouter | $1.4 | $4.4 | Serverless |
| GLM-5.3 | Featherless | $1.4 | $4.4 | Serverless |
Popular comparisons in this family
Comparisons
- DeepSeek V4 Pro vs GLM-5.1
- DeepSeek V4 Flash vs GLM-5.1
- GLM-5.2 vs GLM-5.1
- GLM-5.2 vs DeepSeek V4 Pro
- GLM-5.2 vs DeepSeek V4 Flash
- GLM-5.2 vs Composer 2.5
- GLM-5.2 vs Claude Opus 4.8
- GLM-5.2 vs GPT-5.3-Codex
Models(9)
GLM-5.3-Flash
GLM-5.3
GLM-5.2
GLM-5.1
GLM-5V-Turbo
GLM-5 Turbo
GLM-5 9B
GLM-5
Zhipu GLM-5