选择一个模型后会直接打开对应的对比页面。
GLM-5.2 页面汇总基准测试、API 价格和服务商数据，覆盖 306 家服务商，价格从 $0.0010/request 起。GLM-5.2 免费 API 额度来自 7 家服务商。页面同时展示实测速度和首字延迟。
Zhipu GLM-5.2 is Zhipu latest flagship coding and agentic model with a 1M-token context window, enhanced reasoning modes, and long-horizon software engineering capabilities.
基于独立 benchmark family 的实测能力估计，不确定性单独展示。
Input and output token limits for this model, plus how it ranks on long-context understanding.
1 个指标
2 个指标
2 个指标
15 个指标 · 正式评分
11 个指标 · 正式评分
4 个指标 · 估算
10 个指标 · 暂定评分
5 个指标 · 估算
0 个指标 · 暂无数据
1 个指标 · 暂无数据
1 个指标 · 暂定评分
来自 OpenRouter 的第三方端点数据，和 LMSpeed 自测数据分开展示。部分 30 分钟实时性能字段需要配置 OpenRouter API key 后同步才会出现。
| Provider endpoint | 输入 | 输出 | 1 天在线率 | 30m 延迟 | 30m 吞吐 | 上下文 / 输出 |
|---|---|---|---|---|---|---|
Alibaba alibaba/fast | $2.31/M | $7.26/M | 100.0% | — | — | 1.0M tokens / 131.1K tokens |
Z.AI z-ai/fp8 | $1.40/M | $4.40/M | 99.9% | — | — | 1.0M tokens / 131.1K tokens |
Cloudflare cloudflare/fast | $2.10/M | $6.60/M | 99.9% | — | — | 262.1K tokens / 262.1K tokens |
Decart decart/fp4 | $0.720/M | $1.80/M | 99.9% | — | — | 1.0M tokens / 1.0M tokens |
Wafer wafer | $1.26/M | $3.96/M | 99.8% | — | — | 1.0M tokens / 131.1K tokens |
Phala phala | $1.40/M | $4.40/M | 99.8% | — | — | 1.0M tokens / 131.1K tokens |
Novita novita/fp8 | $0.717/M | $2.25/M | 99.8% | — | — | 1.0M tokens / 131.1K tokens |
SiliconFlow siliconflow/fp8 | $1.19/M | $3.74/M | 99.8% | — | — | 1.0M tokens / 262.1K tokens |
Wafer wafer/fast | $2.10/M | $6.60/M | 99.8% | — | — | 1.0M tokens / 131.1K tokens |
Cloudflare cloudflare | $1.40/M | $4.40/M | 99.7% | — | — | 262.1K tokens / 262.1K tokens |
StreamLake streamlake/fp8 | $0.420/M | $1.32/M | 99.6% | — | — | 1.0M tokens / 128K tokens |
CoreWeave coreweave/fp4 | $0.760/M | $2.42/M | 99.5% | — | — | 262.1K tokens / 262.1K tokens |
Friendli friendli | $1.40/M | $4.40/M | 99.4% | — | — | 1.0M tokens / 1.0M tokens |
Fireworks fireworks | $1.40/M | $4.40/M | 99.4% | — | — | 1.0M tokens / — |
Parasail parasail/fp4 | $1.40/M | $4.40/M | 99.3% | — | — | 262.1K tokens / 262.1K tokens |
DeepInfra deepinfra/fp4 | $0.750/M | $2.40/M | 99.2% | — | — | 1.0M tokens / 131.1K tokens |
Alibaba alibaba/fp8 | $0.966/M | $3.04/M | 99.2% | — | — | 1.0M tokens / 131.1K tokens |
BaseTen baseten/fast | $2.10/M | $6.60/M | 99.2% | — | — | 524.3K tokens / 262.1K tokens |
GMICloud gmicloud/fp8 | $0.924/M | $2.90/M | 99.1% | — | — | 1.0M tokens / — |
BaseTen baseten/fp8 | $1.40/M | $4.40/M | 99.0% | — | — | 1.0M tokens / 262.1K tokens |
Inceptron inceptron/fp4 | $0.940/M | $2.90/M | 98.9% | — | — | 1.0M tokens / 1.0M tokens |
AtlasCloud atlas-cloud/fp8 | $1.26/M | $3.96/M | 98.7% | — | — | 1.0M tokens / 131.1K tokens |
Venice venice/fp8 | $1.40/M | $4.40/M | 98.6% | — | — | 1M tokens / 131.1K tokens |
Baidu baidu/fp8 | $0.756/M | $2.38/M | 98.6% | — | — | 1.0M tokens / 131.1K tokens |
DigitalOcean digitalocean | $1.05/M | $4.40/M | 98.5% | — | — | 262.1K tokens / — |
Fireworks fireworks/fast | $2.10/M | $6.60/M | 98.3% | — | — | 1.0M tokens / — |
Sail Research sail-research/fp8 | $1/M | $3.50/M | 98.3% | — | — | 1.0M tokens / 131.1K tokens |
Ionstream ionstream/fp4 | $1.40/M | $4.40/M | 97.9% | — | — | 1.0M tokens / 131.1K tokens |
Together together | $1.40/M | $4.40/M | 96.5% | — | — | 512K tokens / — |
Ambient ambient/fp8 | $1.05/M | $4.40/M | 95.5% | — | — | 202.8K tokens / 202.8K tokens |
Morph morph | $1.10/M | $4.10/M | 95.2% | — | — | 1.0M tokens / 1.0M tokens |
Chutes chutes/fp4 | $1.25/M | $3.95/M | 94.1% | — | — | 1.0M tokens / 65.5K tokens |
AkashML akashml/fp8 | $0.770/M | $2.42/M | 88.3% | — | — | 96.9K tokens / 96.9K tokens |
对比 GLM-5.2 在 299 家服务商的 API 价格。价格从 $0.0010/request 到 $37500.00/M，其中 X666 API 提供最低价 $0.0010/request。 7 家服务商提供免费 API 额度或免费套餐。
| 服务商 | 健康度 | 模型变体 | 分组 | 输入 ($/M) | 输出 ($/M) | 速度 (t/s) | 首字延迟 | 检测 |
|---|---|---|---|---|---|---|---|---|
L1 100% | glm-5.2 | glm | -69%$0.438/M Cache read$0.110/M | -65%$1.53/M | 36.3 t/s | 9.60 s | — | |
L1 100% | GLM-5.2-C | glm | $0.0027/request | - | — | — | ||
L1 100% | glm-5.2-20260613 | glm | -69%$0.438/M Cache read$0.110/M | -65%$1.53/M | — | — | — | |
L1 100% | glm-5.2 | 鱼干公益基础组 | -99%$0.020/M Cache read$0.020/M | -99%$0.040/M | 26.2 t/s | 16.25 s | ||
L1 0% | z-ai/glm-5.2 | default | -95%$0.068/M | -98%$0.068/M | 21.8 t/s | 13.00 s | ||
L1 100% | glm-5.2 | default | -61%$0.548/M Cache read$0.137/M | -56%$1.92/M | — | — | — | |
L1 100% | glm-5.2 | glm-sale | -55%$0.629/M Cache read$0.157/M | -50%$2.20/M | — | — | — | |
L1 100% | GLM-5.2 | default | $0.080/request | - | — | — | — | |
L1 100% | glm-5.2 | sale | $4.00/M | $14.00/M | — | — | — | |
L1 99% L2 0% | z-ai/glm-5.2 | default | -10%$1.26/M Cache read$2.00/M | -10%$3.96/M | — | — | — | |
L1 100% | glm-5.2 | default | -38%$0.870/M Cache read$0.156/M | -39%$2.70/M | — | — | ||
L1 100% | ZhipuAI/GLM-5.2 | default | $0.0014/request | - | — | — | — | |
L1 100% | [按次稳定]ZhipuAI/GLM-5.2 | default | $0.0055/request | - | — | — | — | |
L1 100% | glm-5.2 | default | -84%$0.219/M Cache read$0.055/MCache write$0.055/MCache write 1h$0.088/M | -83%$0.767/M | — | — | — | |
L1 100% | glm-5.2 | default | $8.00/M Cache read$2.00/M | $28.00/M | — | — | — | |
L1 100% L2 42% | glm-5.2 | Z-AI | $23.73/M Cache read$1.42/M | $71.17/M | — | — | — | |
L1 100% L2 100% | glm-5.2 | default | $75.00/M | $75.00/M | — | — | — | |
L1 100% | glm-5.2 | 🇨🇳国产官方模型 | $4.80/M Cache read$1.20/M | $16.80/M | — | — | — | |
L1 100% | glm-5.2 | default | $8.00/M Cache read$2.00/M | $28.00/M | — | — | — | |
兔子API 免费 | L1 100% | glm-5.2 | default | 免费 | 免费 | — | — | — |
