素墨API

A non-profit AI infrastructure offering free access to integrated large language models with privacy-focused, no-logging policies.

素墨API offers 788 LLM API models.

API pricing per token ranges from $0.010 to $0.100/M (input).

Speed benchmark average: 358 tok/s.

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

Avg Speed357.92 tok/s

Latency6.95 s

Total Tests196

Models31

Updated4/17/2026

Created At2/8/2026

Supported Models

Model	Speed	Latency	Tests
llama3.1-8B	1421.44 tok/s	0.94s	10
快速/llama3.1-8B	1140.96 tok/s	1.24s	15
gpt-oss-120b	970.77 tok/s	0.94s	5
echo	827.30 tok/s	0.72s	10
llama3.1-8b	672.05 tok/s	1.17s	15
gpt-oss-20b:free	339.45 tok/s	2.51s	5
翻译/glm-4.7	308.58 tok/s	3.72s	5
gpt-oss-120b:free	257.64 tok/s	2.50s	5
openai/gpt-oss-20b	234.37 tok/s	1.48s	15
google/gemma-3-1b-it	176.46 tok/s	0.87s	5
gpt-oss-120b-medium	160.66 tok/s	8.29s	5
gemini-3-flash-preview	146.12 tok/s	8.32s	5
gemini-2.5-flash	136.00 tok/s	11.32s	5
qwen3-1.7b:free	133.77 tok/s	5.08s	5
测试/gemini-2.5-pro	82.44 tok/s	17.04s	5
测试/gemini-3-flash-preview	77.58 tok/s	14.56s	3
mimo-v2-flash	76.18 tok/s	5.01s	5
qwen3-omni-flash-2025-12-01	61.51 tok/s	5.13s	5
minimax-m2.1	60.12 tok/s	4.51s	5
llama-3.1-nemotron-nano-4b-v1.1	59.42 tok/s	3.12s	5

Showing 20 of 31 models.