# GLM-5.3 Usage, Cost & Rank | OpenCode Data

> GLM-5.3 ranked #22 by tokens across OpenCode last week, with 0.2% of tokens over the past two months. GLM-5.3 costs $1.40 per 1M input tokens and $4.40 per 1M output tokens.

Flagship GLM model for long-horizon coding, agents, and complex project delivery

- Page: https://opencode.ai/data/zhipuai/glm-5-3
- JSON: https://opencode.ai/data/zhipuai/glm-5-3.json
- Model ID: zhipuai/glm-5.3
- Lab: Zhipu
- Updated: 2026-10-02T08:39:37.000Z

## Model facts

| Fact | Value |
| --- | --- |
| Context window | 1M |
| Max output | 131K |
| Knowledge cutoff | - |
| Release date | 2026-08-14 |
| Input modalities | text |
| Output modalities | text |
| Reasoning | Yes |
| Tool calling | Yes |
| Open weights | Yes |

## Pricing: USD per 1M tokens

| Input | Output | Cached input | Cache write |
| --- | --- | --- | --- |
| $1.40 | $4.40 | $0.26 | $0.00 |

## OpenCode usage: past 2 months

| Metric | Value |
| --- | --- |
| Rank by tokens last week | #22 |
| Tokens | 1.4T |
| Share of all tokens | 0.2% |
| Change vs previous 2 months | +100% |
| Unique users | 325K |
| Completed sessions | 605,954 |
| Average tokens per session | 2.4M |
| Average cost per session | $0.8978 |
| Total spend | $544,015 |
| Input tokens served from cache | 92.9% |
| Weekly retention | 53.6% |

## Daily usage: past 2 months

| Date | Tokens | Unique users | Sessions |
| --- | --- | --- | --- |
| 2026-08-14 | 39B | 8.2K | 155 |
| 2026-08-15 | 64B | 8.7K | 94 |
| 2026-08-16 | 53B | 9.1K | 303 |
| 2026-08-17 | 83B | 13K | 1,061 |
| 2026-08-18 | 74B | 13K | 12,286 |
| 2026-08-19 | 69B | 15K | 14,040 |
| 2026-08-20 | 71B | 13K | 13,056 |
| 2026-08-21 | 58B | 12K | 14,610 |
| 2026-08-22 | 39B | 9.1K | 10,464 |
| 2026-08-23 | 41B | 7.2K | 8,929 |
| 2026-08-24 | 74B | 11K | 13,864 |
| 2026-08-25 | 58B | 11K | 38,786 |
| 2026-08-26 | 38B | 9.8K | 10,973 |
| 2026-08-27 | 32B | 8.1K | 8,853 |
| 2026-08-28 | 29B | 6.4K | 8,804 |
| 2026-08-29 | 21B | 4.8K | 6,386 |
| 2026-08-30 | 21B | 5.1K | 5,429 |
| 2026-08-31 | 31B | 6.8K | 8,230 |
| 2026-09-01 | 25B | 6.6K | 7,538 |
| 2026-09-02 | 24B | 6.4K | 7,963 |
| 2026-09-03 | 23B | 6.7K | 7,484 |
| 2026-09-04 | 18B | 5.9K | 10,022 |
| 2026-09-05 | 13B | 3.9K | 9,989 |
| 2026-09-06 | 13B | 3.9K | 8,449 |
| 2026-09-07 | 22B | 5.9K | 21,818 |
| 2026-09-08 | 19B | 5.4K | 20,859 |
| 2026-09-09 | 17B | 4.9K | 12,422 |
| 2026-09-10 | 16B | 5.1K | 12,500 |
| 2026-09-11 | 14B | 4.5K | 10,840 |
| 2026-09-12 | 12B | 3.5K | 11,143 |
| 2026-09-13 | 12B | 3.7K | 9,041 |
| 2026-09-14 | 21B | 5.5K | 13,090 |
| 2026-09-15 | 18B | 5.3K | 13,813 |
| 2026-09-16 | 19B | 5.5K | 15,211 |
| 2026-09-17 | 21B | 6K | 17,923 |
| 2026-09-18 | 20B | 5.6K | 18,000 |
| 2026-09-19 | 17B | 4.5K | 17,259 |
| 2026-09-20 | 16B | 4.5K | 16,567 |
| 2026-09-21 | 22B | 6K | 17,799 |
| 2026-09-22 | 20B | 5.6K | 18,664 |
| 2026-09-23 | 17B | 4.7K | 15,955 |
| 2026-09-24 | 16B | 4.3K | 13,989 |
| 2026-09-25 | 13B | 3.6K | 11,090 |
| 2026-09-26 | 11B | 3.2K | 11,112 |
| 2026-09-27 | 15B | 3.4K | 10,584 |
| 2026-09-28 | 17B | 4.9K | 18,876 |
| 2026-09-29 | 17B | 4.9K | 16,559 |
| 2026-09-30 | 17B | 4.6K | 14,730 |
| 2026-10-01 | 15B | 4.1K | 13,336 |
| 2026-10-02 | 4B | 1.5K | 5,006 |

## Top countries: past 2 months

| Rank | Country | Tokens | Share |
| --- | --- | --- | --- |
| 1 | United States | 288B | 20% |
| 2 | China | 229B | 15.9% |
| 3 | Germany | 68B | 4.7% |
| 4 | Brazil | 55B | 3.8% |
| 5 | Japan | 53B | 3.7% |
| 6 | Singapore | 40B | 2.8% |
| 7 | Hong Kong | 39B | 2.7% |
| 8 | India | 39B | 2.7% |
| 9 | Spain | 36B | 2.5% |
| 10 | United Kingdom | 33B | 2.3% |
| 11 | Canada | 31B | 2.2% |
| 12 | France | 31B | 2.2% |
| 13 | Russia | 29B | 2% |
| 14 | Netherlands | 25B | 1.8% |
| 15 | Italy | 22B | 1.5% |

## Nearby models by tokens last week

| Rank | Model | Lab | Tokens |
| --- | --- | --- | --- |
| 18 | [gpt-5.6-luna](https://opencode.ai/data/openai/gpt-5-6-luna.md) | OpenAI | 195B |
| 19 | [minimax-m3](https://opencode.ai/data/minimax/minimax-m3.md) | MiniMax | 181B |
| 20 | [mai-experimental-test](https://opencode.ai/data/unknown/mai-experimental-test.md) | - | 180B |
| 21 | [kimi-k2.7-code](https://opencode.ai/data/moonshotai/kimi-k2-7-code.md) | Moonshot | 136B |
| 22 | [glm-5.3](https://opencode.ai/data/zhipuai/glm-5-3.md) | Zhipu | 95B |
| 23 | [fledge-alpha](https://opencode.ai/data/unknown/fledge-alpha.md) | - | 90B |
| 24 | [qwen3.7-plus](https://opencode.ai/data/alibaba/qwen3-7-plus.md) | Qwen | 76B |
| 25 | [glm-5.2](https://opencode.ai/data/zhipuai/glm-5-2.md) | Zhipu | 61B |
| 26 | [kimi-k3](https://opencode.ai/data/moonshotai/kimi-k3.md) | Moonshot | 45B |
| 27 | [mimo-v2.5-pro](https://opencode.ai/data/xiaomi/mimo-v2-5-pro.md) | Xiaomi | 42B |

## Benchmarks

| Benchmark | Score | Metric | Source |
| --- | --- | --- | --- |
| Terminal-Bench | 88.2 | pass@1 | https://huggingface.co/zai-org/GLM-5.3 |
| Terminal-Bench | 28.3 | avg@3 | https://huggingface.co/zai-org/GLM-5.3 |
| DeepSWE | 66.9 | resolved | https://huggingface.co/zai-org/GLM-5.3 |
| NL2Repo | 58 | score | https://huggingface.co/zai-org/GLM-5.3 |
| ProgramBench | 19 | almost solved | https://huggingface.co/zai-org/GLM-5.3 |
| FrontierSWE | 78.1 | dominance score | https://huggingface.co/zai-org/GLM-5.3 |
| SWE-Marathon | 42.5 | score | https://huggingface.co/zai-org/GLM-5.3 |
| PostTrainBench | 39.8 | weighted average | https://huggingface.co/zai-org/GLM-5.3 |
| CyberGym | 84.5 | pass@1 | https://huggingface.co/zai-org/GLM-5.3 |
| ExploitGym | 105 | tasks solved | https://huggingface.co/zai-org/GLM-5.3 |
| ExploitGym | 130 | tasks solved | https://huggingface.co/zai-org/GLM-5.3 |
| ExploitBench | 54.4 | average coverage | https://huggingface.co/zai-org/GLM-5.3 |
| Toolathlon-Verified | 73 | pass@1 | https://huggingface.co/zai-org/GLM-5.3 |
| AutomationBench | 48.2 | pass@1 | https://huggingface.co/zai-org/GLM-5.3 |
| Agents' Last Exam | 28.5 | score | https://huggingface.co/zai-org/GLM-5.3 |
| Humanity's Last Exam | 62.5 | accuracy | https://huggingface.co/zai-org/GLM-5.3 |
| GDPval-AA | 1769 | Elo | https://huggingface.co/zai-org/GLM-5.3 |

## Methodology

- Updates: Aggregated every hour. Days and weeks use UTC.
- Tokens: Input, output, reasoning, and cached tokens for each request.
- Users and sessions: Approximate counts of distinct users and OpenCode sessions.
- Cost: Session cost is the average cost per OpenCode session. Token prices are list prices from the OpenCode model catalog.
- Retention: The share of a model's users in one week who use it again the next week.
- Citation: Cite OpenCode Data (opencode.ai/data) with the update time shown at the top of the page.