微B API 提供 OpenAI 兼容的 API 中转服务，支持访问 AI 语言模型。
微B API 目前提供 4 个大模型 API。
速度基准测试平均吞吐 273 tok/s。
new.wei.bi| 模型 | 速度 | 延迟 | 测试数 |
|---|---|---|---|
0311codelalama:latest | 624.78 tok/s | 1.58s | 5 |
19.53 tok/s | 29.76s | 5 | |
217.38 tok/s | 0.53s | 4 | |
217.41 tok/s | 0.52s | 5 |
| 时间 | 模型 | 速度 | 延迟 |
|---|---|---|---|
| Feb 24, 10:19 AM | 0311codelalama:latest | 624.78 tok/s | 1.58s |
| Feb 23, 03:55 PM | deepseek-r1:671b | 19.53 tok/s | 29.76s |
| Feb 19, 04:02 PM | deepseek-ai/DeepSeek-R1 | 217.38 tok/s | 0.53s |
| Feb 19, 03:49 PM | deepseek-r1:1.5b | 217.41 tok/s | 0.52s |
dashscope.aliyuncs.com
Alibaba Cloud DashScope provides AI model APIs including Qwen LLMs, vision, audio, and embedding models.
api.lkeap.cloud.tencent.com
Tencent Cloud LKE API provides access to Tencent's Hunyuan series and other LLMs via OpenAI-compatible endpoints.
qianfan.baidubce.com
Baidu Qianfan is Baidu Intelligent Clouds enterprise AI platform, providing access to ERNIE series models and other LLM capabilities through developer APIs.
api.suanli.cn
Provides OpenAI-compatible API access to open-source AI models like DeepSeek and QwQ, powered by shared idle computing resources.
ark.cn-beijing.volces.com
Volcengine Ark is ByteDance's enterprise AI platform, offering Doubao series models and third-party LLMs via OpenAI-compatible API.
api.ppinfra.com
PPIO is a distributed cloud computing provider offering AI inference, model serving, and edge computing infrastructure for developers and enterprises.
排名基于社区提交的测试数据与定期健康探测，仅供参考，非官方数据。
