阿里云百炼

Current Qwen flagship (qwen3.7-max) for reasoning and text—promo until 2026-06-22: halved Credits with implicit cache.

qwen3.7-plus covers reasoning, vision, and text—a cost-effective workhorse below Max for daily engineering and multimodal tasks.

qwen3.6-plus balances reasoning, vision, and text—help docs use it as a Credits example (~3.18 Credits per call); a solid daily default.

qwen3.6-flash prioritizes speed and cost while supporting reasoning, vision, and text—good for frequent short requests and latency-sensitive work.

qwen-image-2.0 is Qwen image generation—shares the Credits pool with text models for agent or design workflows.