Xiao Wan

A free, community-driven API provider offering limited access to large language models like Grok and Gemini, with a focus on experimental and low-cost usage.

Xiao Wan offers 240 LLM API models.

API pricing per token ranges from $0.0010 to $100.00/M (input).

Speed benchmark average: 76 tok/s.

Avg Speed76.20 tok/s

Latency6.38 s

Total Tests10

Models2

Updated4/17/2026

Created At2/17/2026

Website

API Endpoints

web.xiaowan.ggff.net

Model Pricing

Pricing for 240 models (group: default, updated: 4/17/2026)

Model	Input ($/M)	Output ($/M)	Model Ratio	Completion Ratio
gpt-oss-120bFixed	$0.0010	$0.0010	0	0
grok-4.1-miniFixed	$0.010	$0.010	0	0
grok-4.1-fastFixed	$0.050	$0.050	0	0
grok-4.20-betaFixed	$0.100	$0.100	0	0
grok-imagine-1.0-editFixed	$0.100	$0.100	0	0
grok-4.1-expertFixed	$0.100	$0.100	0	0
grok-4.1-thinkingFixed	$0.100	$0.100	0	0
grok-imagine-1.0Fixed	$0.500	$0.500	0	0
Qwen/QVQ-72B-Preview	$75.00	$75.00	37.5	1
Qwen/Qwen3-8B	$75.00	$75.00	37.5	1
LoRA/Qwen/Qwen2.5-72B-Instruct	$75.00	$75.00	37.5	1
gemini-3-flash(low)	$75.00	$300.00	37.5	4
gpt-5.3-codex	$75.00	$600.00	37.5	8
Kwaipilot/KAT-Dev	$75.00	$75.00	37.5	1
Pro/Qwen/Qwen2-7B-Instruct	$75.00	$75.00	37.5	1
deepseek-ai/deepseek-vl2	$75.00	$75.00	37.5	1
gemini-3-pro-high(none)	$75.00	$450.00	37.5	6
black-forest-labs/FLUX.1-schnell	$75.00	$75.00	37.5	1
deepseek-ai/DeepSeek-R1-Distill-Qwen-7B	$75.00	$75.00	37.5	1
deepseek-ai/DeepSeek-V2.5	$75.00	$75.00	37.5	1

Showing 20 of 240 models.