Dream API

A unified LLM API gateway providing access to multiple AI models with enterprise-grade security, low latency, and high concurrency.

Categories

Dream API offers 16 LLM API models.

Speed benchmark average: 93 tok/s.

Avg Speed92.56 tok/s

Latency1.64 s

Total Tests92

Models16

Updated4/18/2026

Created At8/13/2025

Versionv12.1.5

Online SinceApr 16, 2026, 05:30 PM

DrawingTaskData ExportCheck-in

GitHub

Recent Test Records

Time	Model	Speed	Latency
Feb 20, 03:09 AM	gpt-4o	86.05 tok/s	0.58s
Feb 19, 07:59 AM	deepseek-r1	6.01 tok/s	16.05s
Feb 19, 06:21 AM	gpt-4-0125-preview	28.16 tok/s	0.67s
Feb 19, 06:01 AM	gpt-4o-mini-2024-07-18	130.72 tok/s	0.60s
Feb 19, 05:55 AM	gpt-4-1106-preview	42.07 tok/s	0.82s
Feb 19, 05:53 AM	gpt-3.5-turbo-16k-0613	108.01 tok/s	0.73s
Feb 19, 05:53 AM	gpt-3.5-turbo-16k	145.74 tok/s	0.64s
Feb 19, 05:53 AM	gpt-3.5-turbo-0613	140.29 tok/s	0.50s
Feb 19, 05:52 AM	gpt-3.5-turbo-1106	156.38 tok/s	0.82s
Feb 19, 05:52 AM	gpt-3.5-turbo	121.66 tok/s	0.79s