LogoLMSpeed
  • Home
  • Free
  • Providers
  • Docs
LogoLMSpeed
LogoLMSpeed

The best API speed test tool

GitHubGitHubTwitterX (Twitter)Email
Product
  • Features
  • Pricing
  • FAQ
Leaderboard
  • Overview
  • Speed Ranking
  • Latency Ranking
  • Health Ranking
Models
  • All Models
  • GPT
  • Claude
  • Gemini
  • DeepSeek
  • Llama
  • Qwen
Free Models
  • All Free Models
  • Free GPT
  • Free Claude
  • Free Gemini
  • Free DeepSeek
  • Free Llama
  • Free Qwen
Resources
  • Speed Test
  • Provider Directory
  • Documentation
Legal
  • Cookie Policy
  • Privacy Policy
  • Terms of Service
© 2026 LMSpeed All Rights Reserved.Made by Nexmoe with ❤️
首页交流 QQ 群：1034193296，欢迎中转站站长加入讨论 AI 最热话题、newapi、openclaw 等，获取最新测速动态与反馈支持。
M

Mars HK

Mars HK is an API relay service hosted in Hong Kong, providing access to AI language models through OpenAI-compatible endpoints.

Categories

中转站
OpenAIGPT-5.3 CodexChatGLMGLM-4.7OpenAIGPT-5.4

Mars HK offers 20 LLM API models.

Speed benchmark average: 91 tok/s.

Mars HK interface preview
Avg Speed91.20 tok/s
Latency3.04 s
Total Tests114
Models5
Updated4/16/2026
Created At3/27/2026
Website

API Endpoints

  • mars-hk.duckdns.org:38317
  • mars-hk.duckdns.org:31338

Supported Models

ModelSpeedLatencyTests
OpenAIgpt-5.3-codex-spark
252.12 tok/s
2.20s
14
ChatGLMGLM-4.7
134.12 tok/s
1.16s
15
OpenAIgpt-5.4(xhigh)
85.16 tok/s
5.13s
5
OpenAIgpt-5.4(high)
68.13 tok/s
7.10s
5
OpenAIgpt-5.4
54.52 tok/s
3.16s
75

Available Models

15 models available (last updated: 4/18/2026)

glm-4.7glm-5-turboglm-5.1gpt-5openaigpt-5.1openaigpt-5.1-codexopenaigpt-5.1-codex-maxopenaigpt-5.1-codex-miniopenaigpt-5.2openaigpt-5.2-codexopenaigpt-5.3-codexopenaigpt-5.3-codex-sparkopenaigpt-5.4openaigpt-5.4-miniopenaigpt-5.4-nanoopenai

Leaderboard Rankings

Speed
184.2 tokens/s#10/100
Latency
1.13 s#34/100
OverviewPerformance20PricingTests114HealthEmbed