Alibaba Cloud (Qwen) · Flagship

Qwen3.8-Max API pricing and specs

The first open-weight Qwen Max model, with thinking and non-thinking modes and up to 262K tokens of reasoning.

Input

⁨$1.65⁩

Cached input

⁨$0.21⁩

Output

⁨$4.95⁩

US dollars per million tokens, standard tier, short-prompt rate

Explicit cache reads cost $0.137 per million and cache creation $2.063. The Singapore region charges $2 input and $6 output. Batch pricing is offered only in the Beijing region.

Spec sheet

API model string: qwen3.8-max
Provider: Alibaba Cloud (Qwen)
Context: 1M tokens
Max output: 131K tokens
Released: August 3, 2026
Batch API: No batch discount listed
Documented agent features: Tool callingStructured outputBuilt-in web searchAdjustable reasoningImage input

What a task costs

Four reference workloads priced at this model's list rates, with caching applied after the first call. Your numbers will differ; the calculator lets you change every assumption.

Workload	Per task	Per 1,000 tasks
Support bot 4 calls, short tool lookups	⁨$0.039⁩	⁨$39⁩
Coding agent 30 calls, files and test output	⁨$1.63⁩	⁨$1,629⁩
Browser agent 25 calls, screenshots and page text	⁨$1.15⁩	⁨$1,153⁩
Research agent 15 calls, long search results	⁨$1.04⁩	⁨$1,037⁩