Qwen3 Max Thinking

Qwen3 Max Thinking is a reasoning model associated with OpenRouter. Qwen3 Max Thinking on this URL is the OpenRouter route (router), not a claim that OpenRouter created the weights. OpenRouter is the recorded vendor. Router and host prices can diverge from the creator's API, and availability can move even when the model name stays the same. The recorded context window is 262,144 tokens, with max output around 32,768 tokens. Qwen3 Max Thinking is marked for tool and function calling, explicit reasoning. Speed tier: slow. Stored use cases: General.

Qwen3 Max Thinking is listed at $0.78 per million input tokens and $3.90 per million output tokens (USD). That quote is the recorded API rate for this page's provider route, not a subscription seat and not a blended monthly bill. A short prompt and a long agent trace will not cost the same amount even when the sticker price stays put.

Recorded access points for Qwen3 Max Thinking: OpenRouter (router, available).

How to use this Qwen3 Max Thinking page: read the price as a per-million-token quote, then open the API price directory if you need other providers for the same family. Open a related model page when you are choosing between sizes or between the creator API and a router. Open the product page when the practical question is a consumer subscription (seat, credits, or included models) rather than metered API spend. Status on this record is available, with a stored release date of 2026-02-09.