By the toolist.ai editors. Researched from public sources, not hands-on tested by toolist.ai. Updated November 3, 2026; prices checked November 3, 2026.
GLM-5.3: price, context window and benchmarks
GLM-5.3 is Z.ai's flagship open-weights model for coding and long-running agents, at $1.40 per million input tokens and $4.40 per million output tokens on Z.ai's API, with a context window of 1 million tokens. The weights are on Hugging Face for self-hosting. Artificial Analysis puts it at 32 on its Intelligence Index at max effort, the API default, against 59 for Claude Opus 5.5.
Best for
- Teams that want to run the model themselves: the weights are on Hugging Face, and SGLang and vLLM are supported
- Long coding sessions at a lower token price: $4.40 per million output tokens against $20 for Claude Opus 5.5
- Prompt-heavy agents, since cached input costs $0.26 per million tokens and cached storage is free for a limited time
Key facts
| Fact | GLM-5.3 | Source |
|---|---|---|
| Provider | Z.ai | Z.ai docs: Pricing |
| API model name | glm-5.3 | Z.ai docs: GLM-5.3 |
| Released | August 18, 2026 | Z.ai docs: New Released |
| Input price | $1.40 per million tokens | Z.ai docs: Pricing |
| Cached input price | $0.26 per million tokens | Z.ai docs: Pricing |
| Output price | $4.40 per million tokens | Z.ai docs: Pricing |
| Context window | 1,000,000 tokens | Z.ai docs: GLM-5.3 |
| Maximum output | 128,000 tokens | Z.ai docs: GLM-5.3 |
| Input and output | Text in; text out | Z.ai docs: GLM-5.3 |
| Reasoning | Always reasons; reasoning cannot be turned off. Effort levels low, high and max; max is the API default | Z.ai docs: GLM-5.3 |
| API and cloud platforms | Open weights on Hugging Face under the GLM-5.3 License, which lets anyone use, change and sell the model. A business that sells model access and has over $10 billion in revenue over 12 months must pass a Z.ai security review before commercial use. Also on the Z.ai API | Hugging Face: GLM-5.3 License |
| In the apps | GLM Coding Plan, from $18 a month, with credit limits every 5 hours and every week. Off-peak use costs half the standard credit rate | Z.ai docs: GLM Coding Plan overview |
Cached input storage is free for a limited time. GLM-5.2 costs the same. The smaller GLM-5.3-Flash costs $0.15 in and $0.50 out, and GLM-5.3-FlashX costs $0.37 in and $1.25 out. Web search costs $0.01 a use.
What GLM-5.3 costs a month
Three typical workloads, priced at the standard API rates above. We calculated these; the provider does not publish them.
| Workload | Tokens a month | Cost a month |
|---|---|---|
| Support chatbot 100,000 replies, each 3,000 input and 400 output tokens, no caching | 300M in, 40M out | $596 |
| Document Q&A 5,000 questions over the same 60,000-token document set (90% cache hits), 800 output tokens each | 300M in, 4M out | $130 |
| Coding agent 200 tasks, each 1.5 million input tokens (90% cache hits) and 50,000 output tokens | 300M in, 10M out | $156 |
Reasoning models bill their thinking as output tokens, so heavy reasoning can raise the output count several times. Cache-write charges are not included.
Benchmarks
| Benchmark | Score | Source |
|---|---|---|
| Artificial Analysis Intelligence Index max effort, the API default | 32.0 | Artificial Analysis: GLM-5.3 (max) |
| Terminal-Bench 4.0 max effort | 42.0% | Artificial Analysis: GLM-5.3 vs Claude Opus 5.5 |
| HLE-Diamond max effort | 18.0% | Artificial Analysis: GLM-5.3 vs Claude Opus 5.5 |
| Terminal-Bench-Science 0.1 max effort | 10.0% | Artificial Analysis: GLM-5.3 vs Claude Opus 5.5 |
Speed: about 85.7 output tokens a second (max effort, Artificial Analysis; Artificial Analysis: GLM-5.3 (max)).
Benchmark and speed data marked as such come from Artificial Analysis.
What changed
GLM-5.3 uses the same base model as GLM-5.2, and Z.ai says every gain comes from post-training. Z.ai reports a 50% gain over GLM-5.2 on Z.ai Code Bench, its own in-house test. (Z.ai docs: GLM-5.3)
Other models
- Gemini 4 Carbon: announced, not on sale yet
- GPT-6.1 Terra: $1 in, $5 out per million tokens
- Gemini 4 Argon: $2 in, $10 out per million tokens
- Muse Wonder: $4 in, $20 out per million tokens
- Claude Haiku 5.5: $0.10 in, $0.50 out per million tokens
- GPT-6.1 Sol: $2 in, $10 out per million tokens
- Claude Sonnet 5.5: $2 in, $10 out per million tokens
- Claude Opus 5.5: $4 in, $20 out per million tokens
- GPT-6 Luna: $0.10 in, $0.50 out per million tokens
- GPT-6 Sol: $2 in, $10 out per million tokens
- GPT-6 Astra: $10 in, $50 out per million tokens
- Claude Fable 5.1: $10 in, $50 out per million tokens
- Claude Opus 5: $5 in, $25 out per million tokens
All prices side by side: AI model API price table. How we research: our method.
Sources
- Z.ai docs: GLM-5.3 (checked 2026-11-03)
- Z.ai docs: Pricing (checked 2026-11-03)
- Z.ai docs: New Released (checked 2026-11-03)
- Z.ai docs: GLM Coding Plan overview (checked 2026-11-03)
- Hugging Face: zai-org/GLM-5.3 (checked 2026-11-03)
- Hugging Face: GLM-5.3 License (checked 2026-11-03)
- Artificial Analysis: GLM-5.3 (max) (checked 2026-11-03)
- Artificial Analysis: GLM-5.3 vs Claude Opus 5.5 (checked 2026-11-03)