NewHow we catch a seller listing the wrong model, and refund you if one does.

What is the cheapest Qwen API?

On November 20, 2026, the lowest-priced Qwen model on MartLLM is Qwen: Qwen3.8 Omni Flash at $0.000146 per million input tokens and $0.000456 per million output tokens, 99.9% below its list price. MartLLM serves 32 Qwen models through one OpenAI-compatible API; the table lists the price of each. Prices come from sellers on the MartLLM market and change during the day. MartLLM adds no fee to them.

Prices as of November 20, 2026, 11:00 UTC. This page is rebuilt every hour.

Against other providers: Cheapest Qwen API: below Alibaba Cloud on 3 of 3 models (November 2026)

ModelInput / 1M tokensOutput / 1M tokensBelow list price
Qwen: Qwen3.8 Omni Flash$0.000146$0.00045699.9%
Qwen3 Coder 30B A3B Instruct$0.000693$0.0027799.4%
Qwen3 32B$0.000792$0.0027799.3%
Qwen 3 235B A22B Instruct 2507$0.000567$0.0034799.5%
Qwen: Qwen3.7 Flash$0.00128$0.0055595.7%
Qwen3 Coder Next$0.00119$0.0079299%
Qwen 3.8 27B$0.000804$0.0092599.6%
Qwen3 Next 80B A3B$0.000861$0.0094799.1%
Qwen 3 Coder 480b$0.00297$0.009999.1%
Qwen3 VL 235B A22B$0.00208$0.018899%
Qwen 3.6 35B A3B$0.00447$0.044795.5%
Qwen3.8 2.4T A95B$0.0126$0.037999.4%
Qwen3.8 Flash$0.015$0.04790%
Qwen3.5 Flash$0.02$0.08
Qwen 2.5 7B$0.0275$0.071445.1%
Qwen 3.5 397B$0.017$0.106297%
Qwen 3.6 27B$0.0135$0.134395.8%
Qwen 3.5 9B$0.055$0.082567%
Qwen 3 235B A22B Thinking 2507$0.091$0.09179.8%
Qwen3.5 Plus$0.052$0.312

MartLLM adds no fee to these prices · every model and price →