A3

A3 provides AI model APIs for developers, offering scalable and accessible AI capabilities.

Categories

中转站

Kimi K2 Instruct Gemini 3 Flash Claude Sonnet 4.5 GPT-5.3 Codex GPT-5.2 GLM-4.7 Flash

A3 offers 7 LLM API models.

Speed benchmark average: 159 tok/s.

A3 is an API aggregator, offering models from multiple vendors.

Avg Speed158.58 tok/s

Latency4.11 s

Total Tests40

Models7

Updated4/16/2026

Created At2/6/2026

Website

API Endpoints

a3.awsl.app

Healthy

Last checked 4/17/2026, 3:23:37 AM

Latency

272ms

Day

100%

Recent availability

Week

100%

Recent availability

Month

100%

Recent availability

Latency

283ms

7d avg: 220ms

Health Check

1000History (1000 pts)

PastNow

Health history

Time	Status	Hourly uptime	Latency
2026-04-17	21/21	100%	344ms
2026-04-16	132/132	100%	272ms
2026-04-15	140/140	100%	281ms
2026-04-14	121/121	100%	227ms
2026-04-13	84/84	100%	234ms
2026-04-12	143/143	100%	218ms
2026-04-11	143/143	100%	193ms
2026-04-10	144/144	100%	92ms
2026-04-09	72/72	100%	99ms

Recent Incidents

No incidents reported recently

Health Check Rules

Check target: the provider's dedicated health check URL, or its website URL as fallback.
Request method: HTTP GET with automatic redirect following.
Healthy criteria: HTTP status code 200–399 is considered healthy; any other status or network error is unhealthy.
Timeout: each check has a 5-second timeout. If the server does not respond within 5 s, the check is marked as failed.
Frequency: health checks run periodically via an automated cron job for all providers.

A3

API Endpoints

Health Check

Health history

Recent Incidents

Health Check Rules

A3

API Endpoints

Leaderboard Rankings

Health Check

Health history

Recent Incidents

Health Check Rules

Leaderboard Rankings