A3

Name: A3
Rating: 5.0 (40 reviews)

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

A3 offers 7 LLM API models.

Speed benchmark average: 149 tok/s.

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

Avg Speed158.58 tok/s

Latency4.11 s

Total Tests40

Models7

Updated4/11/2026

Created At2/6/2026

Recent Test Records

Time	Model	Speed	Latency
Feb 7, 11:34 AM	gpt-5.3-codex	45.96 tok/s	2.04s
Feb 7, 11:32 AM	claude-sonnet-4-5	45.99 tok/s	1.35s
Feb 7, 11:30 AM	kimi-k2-instruct-0905	223.21 tok/s	0.29s
Feb 7, 11:29 AM	kimi-k2-instruct-0905	230.60 tok/s	0.28s
Feb 6, 04:18 PM	groq/compound-mini	477.57 tok/s	0.89s
Feb 6, 03:59 PM	gpt-5.2	44.55 tok/s	1.72s
Feb 6, 03:57 PM	gemini-3-flash-preview	161.96 tok/s	5.57s
Feb 6, 03:53 PM	glm-4.7-flash	38.85 tok/s	20.72s