RelayRank - AI API Relay Rankings, Reviews, and Risk Watch

Submit your relay provider to RelayRank

Let more users discover your service. We run real tests and editorial review.

Submit Listing → Real tests Human review Enter watch board

Secure submission

API keys are encrypted and used only for verification tests

Transparent data

Test results and scoring signals are easy to inspect

Continuous monitoring

Round-the-clock monitoring with timely anomaly alerts

All Relay Provider Rankings ⊙

Updated: 6/19/2026, 11:09:07 AM

Live monitoring30s refresh24h observation window

View full ranking →

Rank	Provider	Score	Availability	Latency	Stability	Pricing	Risk	Models	Action
1	YU YunDou Recommended	96	100.00%	2018ms	96%	93%	Low Risk	Claude	Details
2	LI lingxicode Recommended	95	99.40%	2271ms	95%	92%	Low Risk	Claude	Details
3	TO toproutercn	95	98.90%	2058ms	95%	92%	Watch	Claude	Details
4	SA SAIAi	95	98.50%	2216ms	95%	92%	Watch	ClaudeGPT-4o	Details
5	IN infai	94	98.90%	2383ms	95%	91%	Watch	Claude	Details

Scenario Rankings

Rank providers by model and tool scenarios so you can pick the right relay faster.

View all scenario boards →

Claude Code Relay Rankings

Relay providers optimized for Claude Code workflows

Rank	Provider	Score	Availability	Latency
1	YunDou	96	100.00%	2018ms
2	lingxicode	95	99.40%	2271ms
3	toproutercn	95	98.90%	2058ms
4	SAIAi	95	98.50%	2216ms
5	infai	94	98.90%	2383ms

View full Claude board →

OpenAI / Codex Relay Rankings

Best suited for OpenAI API and Codex scenarios

Rank	Provider	Score	Availability	Latency
1	SAIAi	95	98.50%	2216ms
2	right	94	98.20%	2405ms
3	YunDou	96	100.00%	2018ms
4	lingxicode	95	99.40%	2271ms
5	toproutercn	95	98.90%	2058ms

View full Codex board →

Gemini Relay Rankings

Optimized for Google Gemini model channels

Rank	Provider	Score	Availability	Latency
1	SAIAi	95	98.50%	2216ms
2	right	94	98.20%	2405ms
3	YunDou	96	100.00%	2018ms
4	lingxicode	95	99.40%	2271ms
5	toproutercn	95	98.90%	2058ms

View full Gemini board →

Featured Reviews

View all articles →

No featured articles yet. Reviews, pitfall guides, and weekly reports will appear here after publishing.

Risk Watch

View all risk watch →

No risk watch articles yet. New risk analysis will appear here after publishing.

Why RelayRank

Real probes

Requests are tested from multiple nodes at fixed intervals

Transparent scoring

A five-dimension scoring model with visible methodology

Expert review

Experienced developers add hands-on evaluation and feedback

Risk monitoring

Risk signals are watched continuously and surfaced early

Trusted by developer communities

Claude ClaudeClaude community OpenAI OpenAIOpenAI developers Gemini GeminiGemini developers V2EX V2EXV2EX Juejin JuejinJuejin SSPAI SSPaiSSPai Telegram TelegramTelegram groups GitHub GitHubOpen source developers