Qwen API per-million-token pricing across models

On 2026-10-23, the AI section of DEV Community published a piece titled "Qwen API per-million-token pricing across models". It sets out input, cached-input and output rates per million tokens for every Qwen model, with the author's platform's rates listed alongside Qwen's list prices on the official page. The article notes that output costs roughly 3 to 7 times input depending on the model: three times on Qwen3.7 Max and nearly seven on Qwen3.6 27B. The cheapest of the listed models is Qwen3.7 Plus, at $0.192 per million input tokens and $0.768 per million output tokens. It also compares the cost of a single chat turn and of a long-document request on Qwen3.7 Max, 3.7 Plus and 3.6 27B, explaining that a request's cost equals its token count multiplied by the applicable rates of the model named. Each call is billed per token at three rates per million tokens: input, cached input and output. This remains the only available report on the topic, so the story stands at this publication with no follow-up updates.